ELSEVIER
journal homepage: www.elsevier.com/locate/mce
Meceutse and Celular Endocrinology
Induction of CXCL10 chemokine in adrenocortical cells by stimulation through toll-like receptor 3
Eirik Bratland *,1, Alexander Hellesen 1, Eystein S. Husebye
Section for Endocrinology, Institute of Medicine, University of Bergen, N-5020 Bergen, Norway Department of Medicine, Haukeland University Hospital, N-5021 Bergen, Norway
ARTICLE INFO
Article history: Received 26 January 2012 Received in revised form 7 September 2012 Accepted 8 September 2012 Available online 16 September 2012
Keywords:
Addison’s disease CXC chemokine ligand 10 Interferon-y Interferon-inducible protein 10 Toll-like receptor 3
ABSTRACT
Addison’s disease is a prototypic organ-specific autoimmune disease affecting the adrenal cortex. The CXC chemokine ligand 10 (CXCL10) is expressed early in viral infections, and is produced by primary adrenocortical cells stimulated by certain cytokines. CXCL10 is also elevated in the serum of Addison’s disease patients. We therefore investigated if the viral RNA substitute polyinosine-polycytidylic acid (poly (I:C)) could influence the cytokine induced production of CXCL10 by adrenocortical cells. We found that poly (I:C) could induce CXCL10 in NCI-H295R adrenocortical carcinoma cells, either alone or syner- gistically along with cytokines interferon-y and tumor necrosis factor-a. This effect was found to be med- iated by toll-like receptor 3 and both nuclear factor KB (NFKB) and signal transducer and activator of transcription-1 (STAT1), but not type I interferons, seemed to be involved. We propose that the combina- tion of environmental and endogenous factors presented here, could contribute to the multifactorial pathogenesis of autoimmune Addison’s disease.
@ 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
Autoimmune Addison’s disease (AAD),2 or primary adrenocorti- cal insufficiency, is a classic organ-specific autoimmune disorder resulting from a selective immune response against the hormone- producing cells of the adrenal cortex (Bratland and Husebye, 2011). AAD is lethal if left untreated and hormone replacement therapy is the only current treatment option (Erichsen et al., 2009a). AAD is manageable only to a certain extent because of difficulties in restor-
ing the physiological hormone levels and biorhythm (Lovas and Husebye, 2008). The dominant antigen targeted by autoantibodies and autoreactive T cells from AAD patients is steroid cytochrome P450 21-hydroxylase (21OH), an enzyme involved in the cortisol and aldosterone biosynthetic pathway and exclusively expressed in the adrenal cortex (Bratland et al., 2009b; Rottembourg et al., 2010; Winqvist et al., 1992). Genetic factors associated with the risk of developing AAD have been described, the majority of which encodes proteins involved in antigen presentation (e.g. HLA) or T cell regula- tion (CTLA-4, PTPN22) (Blomhoff et al., 2004; Erichsen et al., 2009b; Skinningsrud et al., 2008, 2011). In spite of the knowledge gained in recent years regarding antigens and genetic associations, the role of any environmental agents or endogenous factors is still poorly characterized. A much more detailed understanding of the patho- physiology is therefore needed in order to identify novel biological markers of AAD risk, and to devise new strategies for treatment and prevention.
The actual role played by the cells of the target organ in initia- tion and propagation of autoimmunity has received little attention. There are indications that target cells are not just passive victims to their own destruction by the immune system, but probably play an active role themselves (Hill et al., 2007). Hormone-producing cells of the adrenal cortex have been shown to express many immuno- logically relevant molecules such as human leucocyte antigen (HLA) class II and most of the toll-like receptors (TLRs) (Kanczkow- ski et al., 2009; Marx et al., 2000; McNicol, 1986). Adrenocortical cells may also secrete a wide range of cytokines and chemokines,
* Corresponding author. Address: Section for Endocrinology, Institute of Medi- cine, Laboratorieblokken, 8 etg., N-5021 Bergen, Norway. Tel .: +47 55973077; fax: +47 55972950.
E-mail address: eirik.bratland@med.uib.no (E. Bratland).
1 These authors contributed equally to the manuscript.
2 Abbreviations: 21OH, steroid cytochrome P450 21-hydroxylase; AAD, autoim- mune Addison’s disease; BSA, bovine serum albumin; CXCL9, CXC chemokine ligand 9; CXCL10, CXC chemokine ligand 10; CXCL11, CXC chemokine ligand 11; DAPI, 2-(4- amidinophenyl)-1H-indole-6-carboxamidine; dsRNA, double stranded RNA; FBS, foetal bovine serum; I-TAC, interferon-inducible T cell alpha chemoattractant; HLA, human leukocyte antigen; IFN, interferon; IFNAR2, chain 2 of the interferon-Alpha/ Beta receptor; IL, interleukin; IP-10, interferon-inducible protein-10; IRF3, interferon regulatory factor 3; ISRE, interferon stimulated response element; MAP, mitogen activated protein; MDA5, melanoma-differentiation-associated gene 5; MIG, mono- kine induced by gamma-interferon; NFKB, nuclear factor kappa-light-chain-enhancer of activated B cells; PAMPs, pathogen-associated molecular patterns; PBMCs, peripheral blood mononuclear cells; poly (I:C), polyinosine-polycytidylic acid; RIG- 1, retinoic-acid-inducible protein 1; RIP1, receptor-interacting protein 1; STAT1, signal transducer and activator of transcription 1; TLRs, toll-like receptors; TNF, tumor necrosis factor; TRIF, TIR-domain-containing adapter-inducing interferon B.
including interleukin (IL)-1, IL-6, IL-8, IL-18 and tumor necrosis factor (TNF)-a, under certain conditions (reviewed in Bornstein et al. (2004)). A specific chemokine implicated in AAD and several other autoimmune endocrine disorders is the CXC Chemokine Li- gand 10 (CXCL10, also known as interferon-inducible protein 10, IP10) (Rotondi et al., 2007). This particular chemokine is secreted by primary adrenal zona fasciculata cells after exposure to the inflammatory cytokines TNFx and interferon (IFN)-y, and serum levels of CXCL10 are also elevated in AAD patients compared to controls (Rotondi et al., 2005). CXCL10, in addition to CXCL9 (also known as monokine induced by gamma-interferon, MIG) and CXCL11 (also known as interferon-inducible T cell alpha chemoat- tractant, I-TAC), are ligands for the chemokine receptor CXCR3. CXCR3 is predominantly expressed on activated T cells, while all three CXCR3 ligands can be induced in a wide variety of cells by T cell derived IFNy (Rotondi et al., 2007). An amplification loop can be envisioned between autoreactive T cells that produce IFNy and target tissue cells wherein T cell-derived IFNy induces the pro- duction of CXCR3-binding chemokines, thereby attracting more CXCR3-expressing activated T cells and thus amplifying an ongoing T cell response.
During inflammatory conditions, such as a viral infection, the adrenal cortex may be exposed to both cytokines and microbial products. It is our hypothesis that such conditions may serve to ini- tiate or perpetuate autoimmunity to the adrenal cortex in geneti- cally susceptible individuals and that the activation of the CXCR3/CXCL10 circuit may be one of the mechanisms at play. Sev- eral viral strains, such as members of the Herpesviridae family, are known to cause both acute and latent infections of the adrenal gland and may stimulate the production of CXCL10 trough TLRs (Paolo and Nosanchuk, 2006). Specifically, the presence of viral infection may be sensed by TLR3 which recognizes double stranded RNA (dsRNA), a molecular pattern produced by many viruses dur- ing their infection cycle (Alexopoulou et al., 2001). In order to investigate whether TLR3 could induce CXCL10 production in adre- nocortical cells, either alone or in combination with IFNy or TNFa, we stimulated NCI-H295R adrenocortical carcinoma cells with poly (I:C), a well-characterized TLR3 agonist that may mimic per- sistent subclinical viral infections (Caskey et al., 2011).
2. Materials and methods
2.1. Reagents, antibodies and cell lines
The TLR3 agonist poly (I:C) was purchased from Sigma-Aldrich. Chloroquine, an inhibitor of endosomal acidification, was pur- chased from Invivogen. Recombinant carrier-free human IFNy was obtained from Biolegend. Recombinant carrier-free human TNFa was from Peprotech. Recombinant 21OH was expressed in baculovirus-infected Sf9 Spodoptera frugiperda insect cells, and purified as described (Bratland et al., 2009a,b). A goat polyclonal antibody against human TLR3 (C-20) was obtained from Santa Cruz Biotechnology Inc., Alexa Fluor® 594-conjugated donkey anti-goat IgG was from Invitrogen. A neutralizing monoclonal antibody against TLR3 (mouse anti-human TLR3, clone 3.7, from now on des- ignated TLR3.7 to distinguish it from the antibody used for immu- nofluorescence) was purchased from Enzo Life Sciences. An antagonistic antibody against chain 2 of the type I interferon recep- tor, human interferon-alpha/beta receptor (anti-IFNAR2, clone MMHAR-2), was obtained from PBL interferon source. A cortisol enzyme immunoassay (EIA) kit was purchased from Cayman Chemical. BAY 11-7082, a specific inhibitor of nuclear factor kap- pa-light-chain-enhancer of activated B cells (NFKB), and S14-95, a specific inhibitor of signal transducer and activator of transcrip- tion (STAT)-1, were both from Enzo Life Sciences. All other chemi-
cals and reagents were from Sigma-Aldrich. The human adrenocortical carcinoma cell line NCI-H295R, which expresses all major adrenocortical enzyme systems and their corresponding steroid hormones (Gazdar et al., 1990), was a kind gift from Profes- sor Marit Bakke (Institute of Biomedicine, University of Bergen, Norway). SBAC cells, which are derived from the adrenal cortex of normal cows (Simonian et al., 1982), were purchased from the European Collection of Cell Cultures (ECACCs).
2.2. Patients and controls
Heparinized blood samples were collected from 22 patients with verified autoimmune Addison’s disease and 20 age-matched blood donors. Plasma was aliquoted and stored at -20 ℃ until the time of analysis. Peripheral blood mononuclear cells (PBMCs) were isolated from 10 patients and 10 blood donors by density gra- dient centrifugation using Lymphoprep (Axxis-Shield), and cryo- preserved in human male AB serum (Lonza) containing 10% (v/v) dimethylsulfoxide. All plasma samples were tested undiluted for CXCL9, CXCL10 and CXCL11 content. All patients and blood donors signed informed consent approved by the Health Region West eth- ics committee (149/96-47.96). All experiments were conducted in accordance with the Declaration of Helsinki.
2.3. Cell culture
NCI-H295R cells were grown in DMEM-F12 medium (Sigma- Aldrich) supplemented with 1% insulin, transferrin, selenium (ITS, BD Biosciences), 2.5% NuSerum IV (BD Biosciences) and 100 U/mL penicillin/100 µg/mL streptomycin (Lonza) at 37 ℃ with 5% CO2 in a humidified incubator. For stimulation experiments, 3 x 105 cells were seeded in 24-well culture plates in supplemented med- ium and allowed to grow over night. Following overnight culturing, the medium was removed and the cells washed once with sterile PBS before incubation with different stimuli (poly (I:C) alone or in combinations with cytokines, inhibitors and/or antibodies) in 0.5 or 1 mL supplemented medium. Unless otherwise stated or indicated in figures, the following concentrations were used: poly (I:C), 100 µg/mL; IFNy, 1 µg/mL; TNFx, 100 ng/mL. After 24 h stim- ulation, culture supernatants were harvested and frozen at -80 ℃ until assayed for CXCL10 content.
SBAC cells were grown in Hams F12 medium (Sigma-Aldrich) supplemented with 50 ng/ml fibroblast growth factor and 10% FBS. For stimulation experiments, the cells were treated exactly as NCI-H295R, except that IFNy was omitted due to poor bioactiv- ity of the human cytokine on bovine cells (Pestka et al., 1987).
PBMCs were cultured at 2 × 106 cells/mL in RPMI 1640 supple- mented with 10% (v/v) human male AB serum, 2 mM L-glutamine, 10 mM HEPES buffer, 1 mM sodium pyruvate, 1% (v/v) non-essen- tial amino acids, 100 U/mL penicillin, 100 µg/mL streptomycin and 50 µM 2-mercaptoethanol. Cells were stimulated with 2.5 µg/mL 210H for 72 h and culture supernatants were assayed for IFNy as described previously (Bratland et al., 2009b), using a sandwich ELI- SA kit from Biolegend.
2.4. Immunofluorescence
NCI-H295R or SBAC cells were cultured on 18 mm glass cover- slips in 6-well plates in their respective supplemented media. After 24 h, the coverslips were washed once in PBS and fixed in PBS con- taining 4% (v/v) paraformaldehyde. After another wash in PBS con- taining 3% (w/v) bovine serum albumin (BSA), cells were permeabilized with PBS containing 0.5% (v/v) Triton X. The cells were then blocked for 1 h in PBS with 10% (v/v) foetal bovine ser- um (FBS). Following blocking, anti-TLR3 antibodies diluted 1:100 in PBS with 1% (w/v) BSA were applied for 1 h at room tempera-
ture. After three washes in PBS with 1% BSA, fluorochrome-conju- gated secondary antibody was applied at 1:1000 dilution in PBS with 1% BSA for 1 h at room temperature. Coverslips were subse- quently washed thrice in PBS and once in ddH2O, and mounted onto SuperFrost microscope slides with ProLong Gold anti-fade re- agent with 2-(4-amidinophenyl)-1H-indole-6-carboxamidine (DAPI, Molecular Probes, Invitrogen). Finally, slides were analyzed under a Nikon TE2000 widefield fluorescence microscope with a 60× Plan Aco DIC N2 objective. Images were acquired with a Nikon DS-U2/L2 camera controlled by NIS-Elements AR 3.10 software. The imaging was performed at the Molecular Imaging Center (Fuge, Norwegian Research Council), University of Bergen.
2.5. Chemokine immunoassays
Plasma levels of CXCL9 and CXCL10 were determined by sand- wich ELISA kits from RayBiotech, Inc., Plasma CXCL11 levels were measured by a CXCL11/I-TAC Quantikine ELISA kit from R&D sys- tems. CXCL10 secretion by NCI-H295R cells was measured in cell-culture supernatants by a CXCL10/IP10 ELISA kit from R&D systems. CXCL10 secretion from SBAC cells was analyzed in a com- petitive sandwich ELISA as follows: Nunc-Immuno™ 96-well flat- bottomed plates with MaxiSorp™ surface (Thermo Fisher Scien- tific) were coated with a mouse anti-human CXCL10 capture anti- body (R&D systems, clone 33036) over night. After blocking for 1 h with 1% (w/v) BSA in PBS, recombinant human CXCL10 (R&D sys- tems) was added at 500 pg/mL for 2 h at room temperature. Then, a biotinylated goat anti-human CXCL10 detection antibody, prein- cubated with supernatants from poly (I:C)- or TNFa-stimulated SBAC cells, was added for 2 h at room temperature. Supernatants from untreated SBAC cells served as controls. Finally, after the addition of Streptavidin-HRP, the plates were developed using tet- ramethylbenzidine and H2O2. Reactions were stopped by adding H2SO4 and the amount of CXCL10 in the supernatants was esti- mated based on the decrease in absorbance at 450 nm compared to wells with supernatants from unstimulated SBAC cells.
2.6. Statistics and presentation of data
Quantitative data from patient-based experiments are ex- pressed as means of duplicates or triplicates. All data from experi- ments using the NCI-H295R cell-line are expressed as means of at least three independent experiments. Comparisons of experimental means were tested by two-sided Student’s t-test or non-parametric Mann-Whitney U-test. Welch’s correction was applied when vari- ances of two groups were significantly different (as determined by F-test). Spearman’s rank correlation coefficient was used to test for correlation between two variables. To test for synergism be- tween two variables, a two-way analysis of variance (ANOVA) was performed as recommended by Slinker (1998). For all statisti- cal tests P < 0.05 was considered significant. All statistical opera- tions were performed using GraphPad Prism v 5.02.
3. Results
3.1. Plasma levels of CXCR3 ligands
Initially, we determined plasma levels of all three CXCR3 bind- ing chemokines in AAD patients and age-matched controls (Fig. 1). We confirmed previous reports that levels of CXCL10 are signifi- cantly elevated in AAD patients (mean 212 pg/mL, range 24- 2332 pg/mL) compared to healthy controls (mean 51 pg/mL, range 17-130 pg/mL) (P < 0.01, Fig. 1b). Two patients had very high levels of CXCL10 in their plasma at 2332 and 520 pg/mL, respectively. When these two samples were omitted from the statistical analy-
sis, the difference was still significant (P < 0.01). Levels of CXCL9 were also significantly increased in AAD patients (mean 2123 pg/ mL, range 328-6280 pg/mL) compared to healthy controls (mean 1157, range 212-3953 pg/mL), but not to the same degree as CXCL10 (P < 0.05, Fig. 1a). Patient CXCL11 levels were not signifi- cantly different from controls (Fig. 1c). In vitro production of IFNY by PBMCs in response to 21OH stimulation was significantly ele- vated among AAD patients (mean 305.8 pg/mL, range 4.7-971 pg/ mL) compared to healthy controls (mean 26.0 pg/mL, range 0.0- 79.0 pg/mL) (P < 0.01, data not shown). We also compared plasma levels of the different CXCR3 ligands with 21OH-induced IFNy pro- duction for individual patients in order to investigate if any corre- lation existed. Out of the three CXCR3-binding chemokines, only CXCL10 levels correlated significantly with 21OH-stimulated IFNy production (P < 0.05, Fig. 1d).
3.2. CXCL10 production in NCI-H295R cells
As CXCL10 production previously has been demonstrated in pri- mary human adrenal zona fasciculata cells upon stimulation with IFNy and/or TNFo, we wanted to assess the ability of NCI-H295R adrenocortical carcinoma cells to secrete CXCL10 under similar conditions. CXCL10 could be detected in a dose-dependent manner after stimulation with IFNy or TNFa alone. However, a strong syn- ergistic effect was observed when these two cytokines were added together at both high and low concentrations (P< 0.001 and P < 0.01, respectively, Fig. 2).
3.3. Expression of TLR3 in NCI-H295R cells and the effect of poly (I:C) on CXCL10 production
Immunofluorescence analyses clearly showed the expression of TLR3 in NCI-H295R cells (Fig. 3). TLR3 stained as cytoplasmic punc- tuate dot-like structures, indicating localization in intracellular vesicles.
After confirming the expression of TLR3 in adrenocortical cells, we stimulated NCI-H295R cells with increasing concentrations of the TLR3 agonist poly (I:C), both alone and in combination with IFNy and TNFo. Poly (I:C) alone induced low but significant levels of CXCL10 at high concentrations (P < 0.05, at 100 µg/mL), while the addition of IFNy or TNFa increased levels of CXCL10 substan- tially (Fig. 4). The synergistic effects of poly (I:C) plus IFNy or poly (I:C) plus TNFa were statistically significant (P < 0.01 and P < 0.05, respectively). Neither of poly (I:C), IFNy nor TNFa, alone or in com- bination, had any influence on the cortisol production of the NCI- H295R cells. After 24 h of culture, 250 ± 50 pg/mL of cortisol was detected, regardless of the presence of poly (I:C) or cytokines (data not shown).
3.4. The importance of TLR3 in poly (I:C)-induced CXCL10 production
As several cytosolic sensors for dsRNA have been described, we wanted to determine the relative involvement of TLR3 compared to other poly (I:C)-recognizing receptors in the induction of CXCL10 in NCI-H295R cells. Initially, we added an inhibitor of endosomal acidification, chloroquine, to NCI-H295R cells prior to stimulation with poly (I:C) and cytokines (Fig. 5a). The effect of poly (I:C) alone on CXCL10 production was greatly reduced in the presence of chlo- roquine (P < 0.01). Moreover, the CXCL10 production induced by poly (I:C) plus TNFa was reduced by more than 97% by the addition of chloroquine (P < 0.05). CXCL10 production in the presence of TNFa alone was also inhibited by chloroquine (approximately 60%, P < 0.001). The effect of chloroquine on poly (I:C) plus IFNY or IFNy alone did not reach statistical significance, although in the latter case a trend towards increased CXCL10 production was noted.
a
8000
b
**
*
2500
2000
CXCL9 pg/ml
6000
..
CXCL10 pg/ml
1500
1000
4000
··
..
500
300
2000
200
100
0
AAD patients
Healthy controls
0
AAD patients
Healthy controls
c
800
n.s
d 1500
* P < 0.05
CXCL11 pg/ml
600
IFN-y pg/ml
1000
400
200
500
0
AAD patients
Healthy controls
0
0
200
400
600
CXCL10 pg/ml
3000
**
CXCL10 (pg/mL)
2500
2000
300
200
100
0
Blank
IFNY [0.1]
TNFa [1]
IFNY [0.1] + TNFa [1]
IFNY [1]
TNFa [100]
IFNY [1] + TNFa [100]
To specifically assess the involvement of TLR3, we also incu- bated NCI-H295R cells with a neutralizing antibody against TLR3 (TLR3.7) prior to stimulation with poly (I:C) (Fig. 5b). When stim-
ulating with poly (I:C) alone CXCL10 production was reduced by 70% in the presence of TLR3.7 (P < 0.01), while CXCL10 induced by poly (I:C) plus TNFa was inhibited by 50% (P < 0.05). The effect of TLR3.7 on poly (I:C) plus IFNy-induced CXCL10 production was not statistically significant.
3.5. Role of type I interferons in poly (I:C)- and cytokine-induced CXCL10 production
In order to investigate if the CXCL10 production actually was a secondary effect of a rapid type I interferon (IFNa or IFNß) response after initial poly (I:C) stimulation, we incubated NCI-H295R cells with a neutralizing antibody against the interferon-alpha/beta receptor (IFNAR) along with poly (I:C) and cytokines (Fig. 5c). No effects of IFNAR blockage on CXCL10 production were observed for poly (I:C) alone, or for any combinations of stimuli.
3.6. Effect of NFKB and STAT-1 inhibitors on CXCL10 induction
In order to analyze the different pathways involved, and to gain insight into the mechanisms behind the synergistic effects on CXCL10 production, we added specific inhibitors to NFKB and STAT-1 to different stimulation cultures. Addition of the BAY 11- 7082 NFKB inhibitor, reduced the production of CXCL10 induced by poly (I:C) alone and poly (I:C) plus TNFx by more than 80% or 50%, respectively (P<0.001 and P<0.01, Fig. 6). Likewise, the STAT1 inhibitor S14-95 reduced CXCL10 production by poly (I:C) and poly (I:C) plus IFNy by 85% and 51%, respectively (P < 0.05 and P < 0.01, Fig. 6).
a
b
a
3000
**
2500
CXCL10 (pg/mL)
2000
1500
1000
200
150
100
*
50
0
Poly (I:C)
0
1
10
100
0
1
10
100
ug/ml
IFNY
+
+
+
+
b
1500
*
1200
CXCL10 (pg/mL)
900
600
300
150
100
*
50
0
Poly (I:C) TNFa
0
1
10
100
0
1
10
100
ug/ml
-
+
+
+
+
3.7. Expression and biological activity of TLR3 in normal bovine adrenocortical cells
To assess that the observed CXCL10 responses also could be in- duced in normal adrenocortical cells, SBAC cells were analyzed for the expression of TLR3 using immunofluorescence as described for NCI-H295R cells. Fig. 7 shows that TLR3 is clearly expressed in these cells as well, and that the intracellular distribution seems to be similar to the NCI-H295R cells. Next, we assessed the ability of SBAC cells to produce CXCL10 when stimulated with poly (I:C), TNFa or the combination of these. CXCL10 was produced at signif- icant levels when stimulated with poly (I:C) alone (P < 0.01, Fig. 8).
However, the addition of human TNFa had very little effect, and although co-stimulation with poly (I:C) plus TNFa increased the CXCL10 production slightly compared to poly (I:C) alone, no syner- gistic effect of co-stimulation was evident (results not shown).
4. Discussion
Like most organ-specific autoimmune diseases, susceptibility to AAD is assumed to be determined by a number of factors, including genetic, environmental and endogenous components (Bratland and Husebye, 2011). At present, the best characterized factors are by far the genetic ones, with specific alleles of the HLA complex con- ferring the highest risk (Erichsen et al., 2009b). Little is known about environmental factors, although the adrenal cortex has been described to be permissive to several viral, fungal and bacterial agents (Kelestimur, 2004; McLeod et al., 2011; Trevisan et al., 2009). Regarding endogenous factors, it has become increasingly recognized that adrenocortical cells have a wide range of immuno- logical mediators in their repertoire, including HLA class II, TLRs, cytokines and chemokines. We believe that the expression of these highly immunologically relevant molecules have implications for the immunopathogenesis of AAD. A subclinical or latent viral infec- tion of the adrenal cortex could be envisioned as a possible link be- tween the environmental and the endogenous factors, initiating an autoimmune attack or perhaps perpetuating an already on-going destructive immune response against the adrenal cortex. We therefore hypothesized that viral pathogen-associated molecular patterns (PAMPs) could induce adrenocortical production of CXCL10, a chemokine known to be up-regulated in the circulation of AAD patients.
We initially reproduced earlier findings showing that CXCL10 is elevated in sera of AAD patients compared to healthy controls (Bel- lastella et al., 2011; Kisand et al., 2008; Rotondi et al., 2005). We also determined the levels of related chemokines binding to the same receptor as CXCL10: CXCL9 and CXCL11. Among these, only CXCL9 was significantly up-regulated in AAD plasma compared to healthy controls, although not to the same extent as CXCL10. Moreover, plasma levels of CXCL10 also correlated with T cell pro- duction of IFNy after stimulation with 21OH, the major autoanti- gen in AAD. This could suggest a self-amplifying circuit, where autoreactive T cells directed against 21OH attack the adrenal cor- tex, and serve to induce or enhance adrenocortical secretion of CXCL10 through IFNy production. We therefore focused on CXCL10 as the predominant CXCR3 ligand in the following work.
We were able to demonstrate that NCI-H295R cells could be in- duced to secrete large amounts of CXCL10, similar to what has been previously shown for primary adrenocortical cells (Rotondi et al., 2005). This was evident when either IFNy or TNFa was added to cultures alone, but especially when combined. Compared to the original study by Rotondi et al., we observed a lower production of
a
3500
medium
2500
chloroquine
CXCL10 (pg/mL)
*
1500
500
200
150
100
50
T
0
Poly (I:C)
IFNY
TNFa
Poly (I:C) + IFNY
Poly (I:C) + TNFa
b
2500
2000
medium
*
CXCL10 (pg/mL)
1500
TLR3.7
1000
500
**
60
40
20
0
Poly (I:C)
Poly (I:C) + IFNY
Poly (I:C) + TNFa
c
medium
4500
anti-IFNAR2
3500
CXCL10 (pg/mL)
2500
1500
200
150
100
50
0
Poly (I:C)
IFNy + TNFa.
Poly (I:C) + IFNY
Poly (I:C) + TNFa
CXCL10 per cell, although similar concentrations of cytokines were used. As Rotondi et al. used primary adrenal zona fasciculata cells, this could indicate that the capacity to produce CXCL10 is even higher in terminally differentiated cells than in the pluripotent NCI-H295R adrenocortical cell-line that we have used.
The synergistic effects of IFNy and TNFx on CXCL10 production, which have been observed in several cell types, are probably due to
S14-95
**
2500
2000
BAY 11-7082
medium
CXCL10 (pg/mL)
1500
**
inhibitor
1000
500
60
*
40
20
0
Poly (I:C)
Poly (I:C) + TNFa
Poly (I:C)
Poly (I:C) + IFNy
the cooperative action of IFNy-induced STAT-1 and TNFx-induced NFKB transcription factors at the CXCL10 promoter (Antonelli et al., 2011; Clarke et al., 2010; Lombardi et al., 2009). This is in line with earlier studies reporting that both interferon stimulated re- sponse element (ISRE) and KB binding sites (recognized by STAT1 and NFKB, respectively) are required for maximum synergistic ef- fects of these two cytokines on CXCL10 transcription (Ohmori and Hamilton, 1995).
CXCL10 and the CXCR3 binding chemokines are predominantly induced by interferons, but stimulation with microbial compo- nents have also been reported to induce and/or synergize and en- hance their secretion (Kumar et al., 2006; Loos et al., 2006). Such PAMP-induced CXCL10 production has been shown to be partly mediated through certain TLR receptors. TLR3 and TLR9 are two important TLRs in the innate immune defence against viral patho- gens, with double stranded RNA (dsRNA) and unmethylated DNA as ligands, respectively (Alexopoulou et al., 2001; Hemmi et al., 2000). Both of these TLR receptors have also been shown to be ex- pressed at the mRNA level in primary adrenocortical cells and NCI- H295R adrenocortical carcinoma cells (Kanczkowski et al., 2009). As human primary adrenocortical cells were not available for the current study, we chose to focus on TLR3 since TLR9 protein have been shown to only be expressed at very low levels in NCI- H295R adrenocortical carcinoma cells (Tran et al., 2007). The TLR3 ligand poly (I:C), a synthetic dsRNA substitute, have recently been demonstrated to induce innate immune responses similar to a live viral vaccine in humans (Caskey et al., 2011). Poly (I:C) have also been shown to induce cytokines and chemokines (including CXCL10 and other CXCR3 ligands) in diverse cell types such as cor- neal epithelial cells and thyrocytes (Harii et al., 2005; Kumar et al., 2006).
Having verified that NCI-H295R cells indeed expressed TLR3, we stimulated NCI-H295R cells with poly (I:C). On its own, poly (I:C) was able to induce rather low, but still significant levels of CXCL10. However, when used along with either IFNy or TNFa, the produc- tion of CXCL10 was enhanced in a synergistic manner. The fact that poly (I:C) could substitute for either IFNy or TNFx in enhancing the CXCL10 production suggests the activation of a third signaling
a
b
100
**
CXCL10 (pg/mL)
75
50
25
0
Blank
Poly (I:C)
pathway in the presence of this ligand, possibly involving the inter- feron regulatory factor 3 (IRF3) (Nakaya et al., 2001). However, when poly (I:C) was added in addition to IFNy and TNFa, no further increase in CXCL10 production was observed (data not shown). This could indicate that the CXCL10 producing capacity of the NCI-H295R cell-line is already at its maximum prior to the addition of poly (I:C), or that the effect of poly (I:C) is slightly different along with IFNy or TNFa (discussed below).
Aside from the endosomally located TLR3, dsRNA structures can also be sensed by the cytosolic RNA helicases retinoic-acid-induc- ible protein 1 (RIG-1) and melanoma-differentiation-associated gene 5 (MDA5) (Kato et al., 2006; Malathi et al., 2007). Also, despite detecting a bright immunofluorescent signal for TLR3, we could not reproducibly detect TLR3 using western blot, even when using the same primary anti-TLR3 as with immunofluorescence (data not shown). Therefore, it was important to investigate the specific con- tribution of TLR3 to poly (I:C)-induced CXCL10 expression using chloroquine, which prevents the acidification of endosomes re- quired for the binding of dsRNA to TLR3 (Leonard et al., 2008), and a TLR3 neutralizing antibody (TLR3.7) (Matsumoto et al., 2002). CXCL10 secretion by NCI-H295R cells following poly (I:C) stimulation was reduced by 75% by pre-treatment with chloro- quine, and by close to 70% in the presence of TLR3.7, indicating pre- dominantly TLR3-mediated signaling. The lack of complete blockage by TLR3.7 may be explained by suboptimal antibody con- centrations or a minor role played by the activation of other recep- tors/signaling pathways mentioned above.
In addition to its effect on poly (I:C), chloroquine also affected the production of CXCL10 in IFNy- and TNFx-stimulated cells.
The bisection of TNFa-induced CXCL10 and the trend towards an increased effect of IFNy may be explained by the ability of chloro- quine to both reduce TNF receptor expression and inhibit IFNY receptor 1 chain degradation (Gira et al., 2009; Jeong et al., 2002). This would also account for the effect of chloroquine on the different combinations of poly (I:C), IFNy and TNFa. As for the TLR3.7 experiments, CXCL10 production was less affected by TLR3 blockage in cells stimulated with poly (I:C) plus IFNy com- pared to poly (I:C) plus TNFa. IFNy has previously been shown to upregulate the expression of TLR3 (Ahmad et al., 2010), and could therefore potentially increase the availability of TLR3 receptors in cultures treated with poly (I:C) plus IFNy compared to poly (I:C) plus TNFa, leading to increased TLR3 signaling and CXCL10 production.
Although the induction of CXCL10 by poly (I:C) appears to be TLR3-mediated, the signals leading to CXCL10 transcription could potentially involve an autocrine/paracrine loop with type I inter- ferons as have been reported by others (Doyle et al., 2003; Gautier et al., 2005). To investigate this notion, we employed a neutralizing antibody towards the human interferon alpha/beta receptor chain 2 (anti-IFNAR2). Anti-IFNAR2 blockage gave a maximum of 22% reduction in CXCL10 production, which was observed in cultures treated with poly (I:C) plus TNFo. Thus, a loop involving the initial induction of type I interferons followed by the autocrine/paracrine action of these cytokines does not appear to play a prominent role in CXCL10 induction.
Based on the findings that chloroquine and TLR3.7 could block the CXCL10-inducing effects of poly (I:C), while anti-IFNAR2 could not, we conclude that the enhancing effect of poly (I:C) on CXCL10 secretion is predominantly mediated by TLR3. TLR3 signaling oc- curs through the TIR-domain-containing adapter-inducing inter- feron ß (TRIF) and then diverges into pathways that either activate NFKB or IRF3 (Jiang et al., 2004). The ability of TLR3 to acti- vate both IRF3, which binds to ISRE sites (Reily et al., 2006; Sankar et al., 2006), and NFKB, would explain why the substitution of poly (I:C) for either one of IFNy or TNFa results in a synergistic effect on CXCL10 production as both KB and ISRE sites could be activated in either scenario. While the IFNy and TLR3 signaling pathways ap- pear to operate independently, TNFx and TLR3 signaling converge at the level of receptor-interacting protein 1 (RIP1) (Meylan et al., 2004), ultimately leading to NFKB activation. Thus, increased NFKB activation (still, along with some level of IRF3 activity) would be a likely explanation for the combined effects of poly (I:C) and TNFa. Alternatively, as has been demonstrated for the production of CCL5/RANTES in airway epithelial cells, cross-talk between TLR3 and TNFa may lead to enhanced CXCL10 production via coopera- tive activation of NFKB, while IFNy may stabilize CXCL10 mRNA up-regulated by TLR3 in a STAT-1 dependent manner (Homma et al., 2010). In fact, we showed by using specific inhibitors to NFKB and STAT-1 that both are involved in the induction of CXCL10, even when NCI-H295R cells are treated with poly (I:C) alone. The latter
observation may seem paradoxical, since type I interferons as de- scribed above seemed to play a minor role. However, S14-95, the STAT1 inhibitor we used, is also an inhibitor of the p38 mitogen- activated protein (MAP) kinase, which has been reported to be in- volved in TLR3 mediated CXCL10 production in diverse cell types such as NK cells and keratinocytes (Erkel et al., 2003; Kanda and Watanabe, 2007; Pisegna et al., 2004). Furthermore, STAT1 can also be activated by type III interferons (interferon-) subtypes 1-3) (Dumoutier et al., 2004). Interferon-» subtypes are also induced by poly (I:C), but act through a different receptor than type I inter- ferons in order to activate STAT1, and eventually CXCL10 transcrip- tion (Diegelmann et al., 2010; Pekarek et al., 2007). Interestingly, recent findings suggest that some viral infections in specific tis- sues, e.g. Hepatitis C virus, lead to a rapid type III interferon re- sponse with minimal induction of type I interferons (Thomas et al., 2012).
Finally, we showed that normal bovine adrenocortical cells also expressed TLR3, and that CXCL10 could be induced in a dose dependent manner by poly (I:C) even in these cells. However, TNFx was unable to induce significant amounts of CXCL10, neither alone nor synergistically along with poly (I:C). We observed decreased viability in the SBAC cells after TNFa stimulation, and therefore speculate that the reason for the poor bioactivity of TNFa on these cells could be attributed to an increased sensitivity to TNFa medi- ated apoptosis compared to the NCI-H295R cells. Alternatively, SBAC cells might be less sensitive to the human TNFa used here than bovine TNFa.
5. Conclusion
To conclude, we have demonstrated that adrenocortical cells ex- press a functional TLR3 receptor, and that stimulation through TLR3 may serve to synergistically enhance the production of proinflam- matory chemokine CXCL10. In our view, these observations point at one of the possible mechanisms by which the adrenal cortex may initiate and/or perpetuate the autoimmune attack on itself. However, in addition to increased CXCL10 levels in plasma, a more direct functional link should also be established between the path- ogenesis of AAD and this particular chemokine. Such links have been demonstrated for other autoimmune endocrinopathies, where elevated levels of CXCL10 and infiltration of lymphocytes express- ing the corresponding chemokine receptor CXCR3 were found in the pancreatic lesions of type 1 diabetes patients (Roep et al., 2010). Similarly, adrenocortical cells under attack from the immune system in AAD may be responding to the inflammatory stimuli in- duced by infiltrating lymphocytes and recruit even more immune cells from the circulation by means of chemokine production. Auto- reactive T cells infiltrating the adrenal cortex will perpetuate the production of CXCL10 by secreting IFN-y and other proinflamma- tory cytokines and thus expand the mononuclear infiltrate. Alterna- tively, the production of CXCL10 may be induced in response to a viral infection and in this way contribute to initiate an autoimmune reaction in genetically susceptible individuals harboring anti-adre- nal cortex specific T cells. Future studies should therefore also focus on which viral strains that display tropism towards the adrenal cor- tex, and whether such a viral insult might induce the production of CXCL10 and chemotactic recruitment of T cells.
Acknowledgement
This research is funded by EU FP7, Grant number 201167, Euradrenal.
References
Ahmad, U., Ali, R., Lebastchi, A.H., Qin, L., Lo, S.F., Yakimov, A.O., Khan, S.F., Choy, J.C., Geirsson, A., Pober, J.S., Tellides, G., 2010. IFN-gamma primes intact human coronary arteries and cultured coronary smooth muscle cells to double- stranded RNA- and self-RNA-induced inflammatory responses by upregulating TLR3 and melanoma differentiation-associated gene 5. J. Immunol. 185 (2), 1283-1294.
Alexopoulou, L., Holt, A.C., Medzhitov, R., Flavell, R.A., 2001. Recognition of double- stranded RNA and activation of NF-kappaB by toll-like receptor 3. Nature 413 (6857), 732-738.
Antonelli, A., Ferrari, S.M., Frascerra, S., Corrado, A., Pupilli, C., Bernini, G., Benvenga, S., Ferrannini, E., Fallahi, P., 2011. Peroxisome proliferator-activated receptor alpha agonists modulate Th1 and Th2 chemokine secretion in normal thyrocytes and Graves’ disease. Exp. Cell Res. 317 (11), 1527-1533.
Bellastella, G., Rotondi, M., Pane, E., Costantini, S., Colella, C., Calemma, R., Capone, F., Falorni, A., Castello, G., Sinisi, A.A., Bizzarro, A., Chiovato, L., Bellastella, A., De Bellis, A., 2011. Simultaneous evaluation of the circulating levels of both Th1 and Th2 chemokines in patients with autoimmune Addison’s disease. J. Endocrinol. Invest. 34 (11), 831-834.
Blomhoff, A., Lie, B.A., Myhre, A.G., Kemp, E.H., Weetman, A.P., Akselsen, H.E., Huseby, E.S., Undlien, D.E., 2004. Polymorphisms in the cytotoxic T lymphocyte antigen-4 gene region confer susceptibility to Addison’s disease. J. Clin. Endocrinol. Metab. 89 (7), 3474-3476.
Bornstein, S.R., Rutkowski, H., Vrezas, I., 2004. Cytokines and steroidogenesis. Mol. Cell. Endocrinol. 215 (1-2), 135-141.
Bratland, E., Bredholt, G., Mellgren, G., Knappskog, P.M., Mozes, E., Husebye, E.S., 2009a. The purification and application of biologically active recombinant steroid cytochrome P450 21-hydroxylase: The major autoantigen in autoimmune Addison’s disease. J. Autoimmun. 33 (1), 58-67.
Bratland, E., Husebye, E.S., 2011. Cellular immunity and immunopathology in autoimmune Addison’s disease. Mol. Cell. Endocrinol. 336 (1-2), 180-190.
Bratland, E., Skinningsrud, B., Undlien, D.E., Mozes, E., Husebye, E.S., 2009b. T cell responses to steroid cytochrome P450 21-hydroxylase in patients with autoimmune primary adrenal insufficiency. J. Clin. Endocrinol. Metab. 94 (12), 5117-5124.
Caskey, M., Lefebvre, F., Filali-Mouhim, A., Cameron, M.J., Goulet, J.P., Haddad, E.K., Breton, G., Trumpfheller, C., Pollak, S., Shimeliovich, I., Duque-Alarcon, A., Pan, L., Nelkenbaum, A., Salazar, A.M., Schlesinger, S.J., Steinman, R.M., Sekaly, R.P., 2011. Synthetic double-stranded RNA induces innate immune responses similar to a live viral vaccine in humans. J. Exp. Med. 208 (12), 2357-2366.
Clarke, D.L., Clifford, R.L., Jindarat, S., Proud, D., Pang, L., Belvisi, M., Knox, A.J., 2010. TNFalpha and IFNgamma synergistically enhance transcriptional activation of CXCL10 in human airway smooth muscle cells via STAT-1, NF-kappaB, and the transcriptional coactivator CREB-binding protein. J. Biol. Chem. 285 (38), 29101-29110.
Diegelmann, J., Beigel, F., Zitzmann, K., Kaul, A., Goke, B., Auernhammer, C.J., Bartenschlager, R., Diepolder, H.M., Brand, S., 2010. Comparative analysis of the lambda-interferons IL-28A and IL-29 regarding their transcriptome and their antiviral properties against hepatitis C virus. PLoS ONE 5 (12), e15200.
Doyle, S.E., O’Connell, R., Vaidya, S.A., Chow, E.K., Yee, K., Cheng, G., 2003. Toll-like receptor 3 mediates a more potent antiviral response than Toll-like receptor 4. J. Immunol. 170 (7), 3565-3571.
Dumoutier, L., Tounsi, A., Michiels, T., Sommereyns, C., Kotenko, S.V., Renauld, J.C., 2004. Role of the interleukin (IL)-28 receptor tyrosine residues for antiviral and antiproliferative activity of IL-29/interferon-lambda 1: similarities with type I interferon signaling. J. Biol. Chem. 279 (31), 32269-32274.
Erichsen, M.M., Lovas, K., Fougner, K.J., Svartberg, J., Hauge, E.R., Bollerslev, J., Berg, J.P., Mella, B., Husebye, E.S., 2009a. Normal overall mortality rate in Addison’s disease, but young patients are at risk of premature death. Eur. J. Endocrinol. 160 (2), 233-237.
Erichsen, M.M., Lovas, K., Skinningsrud, B., Wolff, A.B., Undlien, D.E., Svartberg, J., Fougner, K.J., Berg, T.J., Bollerslev, J., Mella, B., Carlson, J.A., Erlich, H., Husebye, E.S., 2009b. Clinical, immunological, and genetic features of autoimmune primary adrenal insufficiency: observations from a Norwegian registry. J. Clin. Endocrinol. Metab. 94 (12), 4882-4890.
Erkel, G., Rether, J., Anke, T., Sterner, O., 2003. S14-95, a novel inhibitor of the JAK/ STAT pathway from a Penicillium species. J. Antibiot. (Tokyo) 56 (4), 337-343.
Gautier, G., Humbert, M., Deauvieau, F., Scuiller, M., Hiscott, J., Bates, E.E., Trinchieri, G., Caux, C., Garrone, P., 2005. A type I interferon autocrine-paracrine loop is involved in toll-like receptor-induced interleukin-12p70 secretion by dendritic cells. J. Exp. Med. 201 (9), 1435-1446.
Gazdar, A.F., Oie, H.K., Shackleton, C.H., Chen, T.R., Triche, T.J., Myers, C.E., Chrousos, G.P., Brennan, M.F., Stein, C.A., La Rocca, R.V., 1990. Establishment and characterization of a human adrenocortical carcinoma cell line that expresses multiple pathways of steroid biosynthesis. Cancer Res. 50 (17), 5488-5496.
Gira, A.K., Kowalczyk, A.P., Feng, Y., Swerlick, R.A., 2009. Iron chelators and hypoxia mimetics inhibit IFNgamma-mediated Jak-STAT signaling. J. Invest. Dermatol. 129 (3), 723-729.
Harii, N., Lewis, C.J., Vasko, V., McCall, K., Benavides-Peralta, U., Sun, X., Ringel, M.D., Saji, M., Giuliani, C., Napolitano, G., Goetz, D.J., Kohn, L.D., 2005. Thyrocytes express a functional toll-like receptor 3: overexpression can be induced by viral infection and reversed by phenylmethimazole and is associated with Hashimoto’s autoimmune thyroiditis. Mol. Endocrinol. 19 (5), 1231-1250.
Hemmi, H., Takeuchi, O., Kawai, T., Kaisho, T., Sato, S., Sanjo, H., Matsumoto, M., Hoshino, K., Wagner, H., Takeda, K., Akira, S., 2000. A toll-like receptor recognizes bacterial DNA. Nature 408 (6813), 740-745.
Hill, N.J., Hultcrantz, M., Sarvetnick, N., Flodstrom-Tullberg, M., 2007. The target tissue in autoimmunity - an influential niche. Eur. J. Immunol. 37 (3), 589-597. Homma, T., Matsukura, S., Hirose, T., Ohnishi, T., Kimura, T., Kurokawa, M., Ieki, K., Odaka, M., Suzuki, S., Watanabe, S., Sato, M., Kawaguchi, M., Schleimer, R.P., Adachi, M., 2010. Cooperative activation of CCL5 expression by TLR3 and tumor necrosis factor-alpha or interferon-gamma through nuclear factor-kappaB or STAT-1 in airway epithelial cells. Int. Arch. Allergy Immunol. 152 (Suppl. 1), 9- 17.
Jeong, J.Y., Choi, J.W., Jeon, K.I., Jue, D.M., 2002. Chloroquine decreases cell-surface expression of tumour necrosis factor receptors in human histiocytic U-937 cells. Immunology 105 (1), 83-91.
Jiang, Z., Mak, T.W., Sen, G., Li, X., 2004. Toll-like receptor 3-mediated activation of NF-kappaB and IRF3 diverges at toll-IL-1 receptor domain-containing adapter inducing IFN-beta. Proc. Natl. Acad. Sci. USA 101 (10), 3533-3538.
Kanczkowski, W., Zacharowski, K., Wirth, M.P., Ehrhart-Bornstein, M., Bornstein, S.R., 2009. Differential expression and action of Toll-like receptors in human adrenocortical cells. Mol. Cell. Endocrinol. 300 (1-2), 57-65.
Kanda, N., Watanabe, S., 2007. Prolactin enhances interferon-gamma-induced production of CXC ligand 9 (CXCL9), CXCL10, and CXCL11 in human keratinocytes. Endocrinology 148 (5), 2317-2325.
Kato, H., Takeuchi, O., Sato, S., Yoneyama, M., Yamamoto, M., Matsui, K., Uematsu, S., Jung, A., Kawai, T., Ishii, K.J., Yamaguchi, O., Otsu, K., Tsujimura, T., Koh, C.S., Reis e Sousa, C., Matsuura, Y., Fujita, T., Akira, S., 2006. Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature 441 (7089), 101-105. Kelestimur, F., 2004. The endocrinology of adrenal tuberculosis: the effects of tuberculosis on the hypothalamicuitary-adrenal axis and adrenocortical function. J. Endocrinol. Invest. 27 (4), 380-386.
Kisand, K., Link, M., Wolff, A.S., Meager, A., Tserel, L., Org, T., Murumagi, A., Uibo, R., Willcox, N., Trebusak Podkrajsek, K., Battelino, T., Lobell, A., Kampe, O., Lima, K., Meloni, A., Ergun-Longmire, B., Maclaren, N.K., Perheentupa, J., Krohn, K.J., Scott, H.S., Husebye, E.S., Peterson, P., 2008. Interferon autoantibodies associated with AIRE deficiency decrease the expression of IFN-stimulated genes. Blood 112 (7), 2657-2666.
Kumar, A., Zhang, J., Yu, F.S., 2006. Toll-like receptor 3 agonist poly(I:C)-induced antiviral response in human corneal epithelial cells. Immunology 117 (1), 11- 21.
Leonard, J.N., Ghirlando, R., Askins, J., Bell, J.K., Margulies, D.H., Davies, D.R., Segal, D.M., 2008. The TLR3 signaling complex forms by cooperative receptor dimerization. Proc. Natl. Acad. Sci. USA 105 (1), 258-263.
Lombardi, A., Cantini, G., Mello, T., Francalanci, M., Gelmini, S., Cosmi, L., Santarlasci, V., Degl’Innocenti, S., Luciani, P., Deledda, C., Annunziato, F., Forti, G., Galli, A., Serio, M., Luconi, M., 2009. Molecular mechanisms underlying the pro- inflammatory synergistic effect of tumor necrosis factor alpha and interferon gamma in human microvascular endothelium. Eur. J. Cell Biol. 88 (12), 731-742.
Loos, T., Dekeyzer, L., Struyf, S., Schutyser, E., Gijsbers, K., Gouwy, M., Fraeyman, A., Put, W., Ronsse, I., Grillet, B., Opdenakker, G., Van Damme, J., Proost, P., 2006. TLR ligands and cytokines induce CXCR3 ligands in endothelial cells: enhanced CXCL9 in autoimmune arthritis. Lab. Invest. 86 (9), 902-916.
Lovas, K., Husebye, E.S., 2008. Replacement therapy for Addison’s disease: recent developments. Exp. Opin. Investig. Drugs 17 (4), 497-509.
Malathi, K., Dong, B., Gale Jr., M., Silverman, R.H., 2007. Small self-RNA generated by RNase L amplifies antiviral innate immunity. Nature 448 (7155), 816-819.
Marx, C., Bornstein, S.R., Wolkersdorfer, G.W., 2000. Cellular immune-endocrine interaction in adrenocortical tissues. Eur. J. Clin. Invest. 30 (Suppl. 3), 1-5.
Matsumoto, M., Kikkawa, S., Kohase, M., Miyake, K., Seya, T., 2002. Establishment of a monoclonal antibody against human Toll-like receptor 3 that blocks double- stranded RNA-mediated signaling. Biochem. Biophys. Res. Commun. 293 (5), 1364-1369.
McLeod, D.S., Mortimer, R.H., Perry-Keene, D.A., Allworth, A., Woods, M.L., Perry- Keene, J., McBride, W.J., Coulter, C., Robson, J.M., 2011. Histoplasmosis in Australia: report of 16 cases and literature review. Medicine (Baltimore) 90 (1), 61-68.
McNicol, A.M., 1986. Class II MHC antigen expression in adrenal cortex. Lancet 2 (8518), 1282.
Meylan, E., Burns, K., Hofmann, K., Blancheteau, V., Martinon, F., Kelliher, M., Tschopp, J., 2004. RIP1 is an essential mediator of Toll-like receptor 3-induced NF-kappa B activation. Nat. Immunol. 5 (5), 503-507.
Nakaya, T., Sato, M., Hata, N., Asagiri, M., Suemori, H., Noguchi, S., Tanaka, N., Taniguchi, T., 2001. Gene induction pathways mediated by distinct IRFs during viral infection. Biochem. Biophys. Res. Commun. 283 (5), 1150-1156.
Ohmori, Y., Hamilton, T.A., 1995. The interferon-stimulated response element and a kappa B site mediate synergistic induction of murine IP-10 gene transcription by IFN-gamma and TNF-alpha. J. Immunol. 154 (10), 5235-5244.
Paolo Jr., W.F., Nosanchuk, J.D., 2006. Adrenal infections. Int. J. Infect. Dis. 10 (5), 343-353.
Pekarek, V., Srinivas, S., Eskdale, J., Gallagher, G., 2007. Interferon lambda-1 (IFN- lambda1/IL-29) induces ELR(-) CXC chemokine mRNA in human peripheral blood mononuclear cells, in an IFN-gamma-independent manner. Genes Immun. 8 (2), 177-180.
Pestka, S., Langer, J.A., Zoon, K.C., Samuel, C.E., 1987. Interferons and their actions. Annu. Rev. Biochem. 56, 727-777.
Pisegna, S., Pirozzi, G., Piccoli, M., Frati, L., Santoni, A., Palmieri, G., 2004. P38 MAPK activation controls the TLR3-mediated up-regulation of cytotoxicity and cytokine production in human NK cells. Blood 104 (13), 4157-4164.
Reily, M.M., Pantoja, C., Hu, X., Chinenov, Y., Rogatsky, I., 2006. The GRIP1:IRF3 interaction as a target for glucocorticoid receptor-mediated immunosuppression. EMBO J. 25 (1), 108-117.
Roep, B.O., Kleijwegt, F.S., van Halteren, A.G., Bonato, V., Boggi, U., Vendrame, F., Marchetti, P., Dotta, F., 2010. Islet inflammation and CXCL10 in recent-onset type 1 diabetes. Clin. Exp. Immunol. 159 (3), 338-343.
Rotondi, M., Chiovato, L., Romagnani, S., Serio, M., Romagnani, P., 2007. Role of chemokines in endocrine autoimmune diseases. Endocrinol. Rev. 28 (5), 492- 520.
Rotondi, M., Falorni, A., De Bellis, A., Laureti, S., Ferruzzi, P., Romagnani, P., Buonamano, A., Lazzeri, E., Crescioli, C., Mannelli, M., Santeusanio, F., Bellastella, A., Serio, M., 2005. Elevated serum interferon-gamma-inducible chemokine-10/ CXC chemokine ligand-10 in autoimmune primary adrenal insufficiency and in vitro expression in human adrenal cells primary cultures after stimulation with proinflammatory cytokines. J. Clin. Endocrinol. Metab. 90 (4), 2357-2363.
Rottembourg, D., Deal, C., Lambert, M., Mallone, R., Carel, J.C., Lacroix, A., Caillat- Zucman, S., le Deist, F., 2010. 21-Hydroxylase epitopes are targeted by CD8 T cells in autoimmune Addison’s disease. J. Autoimmun. 35 (4), 309-315.
Sankar, S., Chan, H., Romanow, W.J., Li, J., Bates, R.J., 2006. IKK-i signals through IRF3 and NFkappaB to mediate the production of inflammatory cytokines. Cell. Signal. 18 (7), 982-993.
Simonian, M.H., White, M.L., Gill, G.N., 1982. Growth and function of cultured bovine adrenocortical cells in a serum-free defined medium. Endocrinology 111 (3), 919-927.
Skinningsrud, B., Husebye, E.S., Gervin, K., Lovas, K., Blomhoff, A., Wolff, A.B., Kemp, E.H., Egeland, T., Undlien, D.E., 2008. Mutation screening of PTPN22: association of the 1858T-allele with Addison’s disease. Eur. J. Hum. Genet. 16 (8), 977-982.
Skinningsrud, B., Lie, B.A., Lavant, E., Carlson, J.A., Erlich, H., Akselsen, H.E., Gervin, K., Wolff, A.B., Erichsen, M.M., Lovas, K., Husebye, E.S., Undlien, D.E., 2011. Multiple loci in the HLA complex are associated with Addison’s disease. J. Clin. Endocrinol. Metab ..
Slinker, B.K., 1998. The statistics of synergism. J. Mol. Cell. Cardiol. 30 (4), 723-731. Thomas, E., Gonzalez, V.D., Li, Q., Modi, A.A., Chen, W., Noureddin, M., Rotman, Y., Liang, T.J., 2012. HCV infection induces a unique hepatic innate immune response associated with robust production of type III interferons. Gastroenterology 142 (4), 978-988.
Tran, N., Koch, A., Berkels, R., Boehm, O., Zacharowski, P.A., Baumgarten, G., Knuefermann, P., Schott, M., Kanczkowski, W., Bornstein, S.R., Lightman, S.L., Zacharowski, K., 2007. Toll-like receptor 9 expression in murine and human adrenal glands and possible implications during inflammation. J. Clin. Endocrinol. Metab. 92 (7), 2773-2783.
Trevisan, M., Matkovic, U., Cusinato, R., Toppo, S., Palu, G., Barzon, L., 2009. Human cytomegalovirus productively infects adrenocortical cells and induces an early cortisol response. J. Cell. Physiol. 221 (3), 629-641.
Winqvist, O., Karlsson, F.A., Kampe, O., 1992. 21-Hydroxylase, a major autoantigen in idiopathic Addison’s disease. Lancet 339 (8809), 1559-1562.