Abstract
Germline mutations in the aryl hydrocarbon receptor-interacting protein (AIP) gene predispose humans to pituitary adenomas through unknown molecular mechanisms. The best-known interacting partner of AIP is the aryl hydrocarbon receptor (AhR), a transcription factor that mediates the effects of xenobiotics implicated in carcinogenesis. As 75% of AIP mutations disrupt the physical and/or functional interaction with AhR, we postulated that the tumorigenic potential of AIP mutations might result from altered AhR signaling. We evaluated the impact of AIP mutations on the AhR signaling pathway, first in fibroblasts from AIP-mutated patients with pituitary adenomas, by comparison with fibroblasts from healthy subjects, then in transfected pituitary GH3 cells. The AIP protein level in mutated fibroblasts was about half of that in cells from healthy subjects, but AhR expression was unaffected. Gene expression analyses showed significant modifications in the expression of the AhR target genes CYP1B1 and AHRR in AIP-mutated fibroblasts, both before and after stimulation with the endogenous AhR ligand kynurenine. Kynurenine increased Cyp1b1 expression to a greater extent in GH3 cells overexpressing wild type compared with cells expressing mutant AIP. Knockdown of endogenous Aip in these cells attenuated Cyp1b1 induction by the AhR ligand. Both mutant AIP expression and knockdown of endogenous Aip affected the kynurenine-dependent GH secretion of GH3 cells. This study of human fibroblasts bearing endogenous heterozygous AIP mutations and transfected pituitary GH3 cells shows that AIP mutations affect the AIP protein level and alter AhR transcriptional activity in a gene- and tissue-dependent manner.
Introduction
Heterozygous germline mutations of the AIP gene predispose humans to pituitary adenomas (Vierimaa et al. 2006). The prevalence of AIP mutations is about 20% in familial isolated pituitary adenomas (Beckers et al. 2013) and 3–4% in sporadic adenomas (Lecoq et al. 2015). AIP is a chaperone protein that interacts with many molecular partners (Trivellin et al. 2011). One partner, the xenobiotic receptor AhR (aryl hydrocarbon receptor), is a ligand-activated transcription factor that mediates adaptive and toxic responses to environmental contaminants such as dioxin (2,3,7,8-tetrachlorodibenzo-p-dioxin, TCDD) (Schmidt et al. 1996). AhR also has physiological functions (Barouki et al. 2007), especially in cell-cycle regulation (Puga et al. 2009), and can play a role in tumorigenesis (Murray et al. 2014). Several endogenous AhR ligands have been identified (Nguyen et al. 2008), including tryptophan metabolites generated by the enzymes IDO (indoleamine 2,3-dioxygenase) and TDO (tryptophan dioxygenase). One such ligand, kynurenine, can promote brain tumor progression through local AhR activation (Opitz et al. 2011). The molecular mechanisms of AIP-dependent pituitary tumorigenesis are elusive. Recent in vitro studies implicate the cAMP pathway, which seems to be activated by AIP deficiency (Formosa et al. 2013, Tuominen et al. 2014), but several lines of evidence also suggest the involvement of AhR. Most pathogenic AIP mutations alter the C-terminal region of the protein, particularly the last few amino acids required for AhR interaction (Bell et al. 2000), resulting in loss of protein–protein interactions (Leontiou et al. 2008, Morgan et al. 2012). AhR is expressed in pituitary cells, and immunohistochemical studies have shown decreased expression of AhR or of ARNT (AhR nuclear translocator), its heterodimerization partner, in human and murine AIP-mutated pituitary adenomas (Heliövaara et al. 2009, Jaffrain-Rea et al. 2009, Raitila et al. 2010). Although only a slight increase in the incidence of pituitary adenomas has been noted since the Seveso accident, which released large amounts of TCDD (Pesatori et al. 2008), an increased prevalence of acromegaly has been observed in other highly polluted industrial areas (Cannavò et al. 2010), suggesting that environmental pollutants can participate in pituitary tumorigenesis, potentially through the AhR pathway.
The impact of AIP mutations on AhR signaling in patients with pituitary adenomas has not yet been investigated. Here, we explored the AhR transcriptional response to its endogenous ligand kynurenine in fibroblasts from patients with pituitary adenomas and heterozygous AIP mutations, and also in the well-characterized somatolactotroph GH3 cell line. We found that AIP mutations resulted in decreased AIP expression and also altered the AhR transcriptional response in both cell models, in a cell- and gene-specific manner. We have further shown that AIP mutants were able to affect kynurenine-dependent GH secretion in GH3 cells.
Materials and methods
Cell lines and treatments
Fibroblasts from four probands (patients P1 to P4) and from four sex- and age-matched controls were obtained by forearm skin biopsy. All the patients and controls gave their written informed consent for cell culture and genetic analysis, and the study was approved by Ethics Committee of Bicêtre Hospital, France. Fibroblasts were cultured in DMEM high-glucose (4.5g/L) medium (PAA, Velizy-Villacoublay, France) supplemented with 20mM HEPES (Life technologies, Saint Aubin, France), antibiotics (penicillin 100IU/mL, streptomycin 100μg/mL), 2mM glutamine (all from PAA), and 10% fetal bovine serum (FBS, Biowest, Nuaillé, France).
GH3 cells (ATCC CCL-82.1, lot 3741980) generously provided by Dr Marie Legendre (U.F. de Génétique Moléculaire, Service de Génétique et d'Embryologie Médicales, Hôpital Trousseau, Paris) were cultured in DMEM/HAM’S F-12 medium (PAA) supplemented with 20mM HEPES, antibiotics (penicillin 100IU/mL, streptomycin 100μg/mL), 2mM glutamine, 2.5% FBS, and 15% donor horse serum (PAA).
All cells were cultured at 37°C in a humidified incubator with 5% CO2. For all the experiments, the cell passage number was between 2 and 20.
Fibroblasts were treated for 8h with 60μM l-kynurenine (Sigma-Aldrich), with or without 10μM 3′,4′-dimethoxyflavone (Sigma-Aldrich) or 0.4μM actinomycin D (Sigma-Aldrich), after 24h of tryptophan deprivation in custom-made DMEM/HAM’S F-12 medium without tryptophan (Life Technologies).
Plasmid constructs
Human AIP cDNA was amplified by PCR from control (wild-type AIP-encoding vector) and proband mRNA (P1, P3, and P4 mutant-encoding vectors) using specific primers (Supplementary Table S1, see section on supplementary data given at the end of this article) in which HindIII and NotI restriction sites were introduced to facilitate cloning into the pcDNATM3.1 plasmid (Life Technologies). All plasmids were sequenced to verify nucleotide identity.
Transfection assay
GH3 cells were transfected with either the empty pcDNATM3.1 vector or with AIP-encoding vectors, using the Lipofectamine 2000 (Life Technologies) reverse-transfection protocol. After 1 day, GH3 cells were deprived of tryptophan for 24h and then exposed to 60μM l-kynurenine for 3h. RNA and protein were extracted 48h post-transfection as described below.
RNA interference
GH3 cells were transfected twice with 100nM rat SMARTpool Aip siRNA or nontargeting siRNA (Dharmacon, GE Healthcare), using Lipofectamine RNAimax (Life Technologies). After 24h of tryptophan deprivation, cells were exposed to l-kynurenine for 3h. RNA and protein were extracted 48h post-transfection as described below.
Reverse transcriptase-PCR (RT-PCR) and quantitative real-time PCR (RT-qPCR)
Total RNA was extracted from cells with the TRI Reagent (Molecular Research Center, Euromedex, Souffelweyersheim, France) according to the manufacturer’s recommendations. After DNAse treatment (Biolabs, Evry, France), RNA was reverse-transcribed with the High-Capacity cDNA Reverse Transcription kit (Life Technologies). Quantitative real-time PCR (RT-qPCR) was performed as described previously (Le Billan et al. 2015). The primer sequences for the genes analyzed by RT-qPCR are shown in Supplementary Table S2.
Western blot (WB)
Total protein extracts were prepared from cells lysed in lysis buffer, 1X Triton X100 and 1X protease inhibitor (Sigma-Aldrich). Immunoblots were incubated for 1h at room temperature (RT) in 5% milk-Tris buffer saline/0.1% Tween 20 before incubation overnight with anti-AIP, anti-β-actin, anti-AhR, anti-α-tubulin, anti-CYP1B1, or anti-GAPDH antibodies (antibody sources and dilutions are shown in the Supplementary Table S3). Blots were then incubated with an IRDye 800-conjugated affinity-purified anti-mouse IgG second antibody (1:15,000 dilution, Perbio Science, Brebières, France) and an IRDye 680-conjugated affinity-purified anti-rabbit IgG second antibody (1:15,000 dilution, Perbio Science) for 1h at RT. Proteins were visualized with an Odyssey-Fc apparatus (LI-COR).
Cell proliferation analysis
Cell proliferation tests were performed using the WST1 assay (Roche) according to the manufacturer’s recommendations. GH3 cells were cultured in 96-well plates, transfected with either the empty pcDNATM3.1 vector or wild-type AIP-encoding vectors using the Lipofectamine 2000 (Life Technologies) reverse-transfection protocol. Cells were deprived of tryptophan and exposed to 60μM l-kynurenine for 48h. Optical densities were measured 4h after the addition of WST1 solution (10μL per well) by spectrophotometry (Viktor, PerkinElmer).
Hormone assays
GH3 cells were cultured in 24-well plates, transfected with either the empty pcDNATM3.1 vector or AIP-encoding vectors using the Lipofectamine 2000 (Life Technologies) reverse-transfection protocol or transfected twice with 100nM rat SMARTpool Aip siRNA or nontargeting siRNA (Dharmacon, GE Healthcare), using Lipofectamine RNAimax (Life Technologies). Cells were deprived of tryptophan and exposed to 60μM l-kynurenine for 48h. After 48h, GH secreted in the medium was determined by EIA as described previously (Steyn et al. 2011), using antibodies listed in the Supplementary Table S3. Values are reported in terms of rGH-RP2. The detection limit was 0.038ng/mL and the intra- and interassay coefficients of variation were 3.2% and below 8.75%, respectively.
Statistical analysis
Results are expressed as mean ± s.e.m. Differences between groups were analyzed by two-way ANOVA followed by a multiple comparisons post-test, or with the nonparametric Mann–Whitney test (Prism software, GraphPad, CA, USA). P values <0.05 were considered statistically significant.
Results
Clinical and genetic characterization of AIP-mutated patients with pituitary adenomas
Table 1 shows the clinical features of the four patients studied here (P1 to P4). As expected in the case of AIP mutation, the patients were young at diagnosis and had macro- or giant, GH-secreting or GH and PRL-cosecreting adenomas. Two of the three patients with somatotropinomas were giants. The AIP mutations were located throughout the coding sequence (Fig. 1A). Patient 1 (M1) had a missense mutation in the start codon that abrogated protein expression. The frameshift mutation (M2) in patient 2 and the nonsense mutation (M4) in patient 4 theoretically resulted in the expression of a protein truncated of the entire C-terminal domain or only the last few amino acids, respectively. The duplication (M3) in patient 3 resulted in a protein with a longer tetratricopeptide repeats (TPR) domain. As AIP is ubiquitously expressed, we used fibroblasts obtained by skin biopsy to study the expression and molecular consequences of the AIP mutants.
AIP expression in fibroblasts and pituitary adenoma of AIP-mutated patients. (A) The four AIP mutations studied here were located throughout the coding sequence and had different predicted effects on the AIP protein. (B) AIP mRNA levels measured by RT-qPCR were lower in the fibroblasts of patients 1 and 2 (M1 and M2) than in controls. AIP transcript levels in M3 and M4 fibroblasts were similar to control values. Results of RT-qPCR are expressed as the mean percentage of control ± s.e.m. of at least six independent determinations performed in duplicate. (C) WB analysis showed lower AIP protein expression in all the patients’ fibroblasts than in controls. (D) Quantitative analysis of six different WB experiments confirmed the decrease in AIP protein expression in mutated fibroblasts. Specific signals for AIP were normalized by the infrared fluorescence of β-actin. Results are shown as the AIP/β-actin ratio and as percentage of control ± s.e.m. (E) WB analysis revealed the absence of AIP protein in the pituitary adenoma of patient 1. For all experiments, values obtained with fibroblasts from the four controls were pooled and arbitrarily set at 100%. *P<0.05, ***P<0.001. M, mutation; bp, base pair; FKBP-PPI, FK506 binding protein-peptidyl-prolyl-isomerase; TPR, tetratricopeptide repeats; AA, amino acid; kDa, kiloDalton.
Citation: Endocrine-Related Cancer 23, 5; 10.1530/ERC-16-0041
Clinical characteristics of the four AIP mutation positive patients with pituitary adenomas.
| Patient | AIP mutation | Sex | Age at diagnosis (years) | Clinical presentation | Adenoma type | IGF-I and/or PRL at diagnosis (ng/mL) | Adenoma size |
|---|---|---|---|---|---|---|---|
| 1 | c.2T>C | F | 42 | Galactorrhea, sweating | GH+PRL | 124/40 | MA |
| p.? | |||||||
| 2 | c.350delG | F | 30 | Primary amenorrhea | PRL | −/7460 | GA |
| p.(Gly117Alafs*39) | |||||||
| 3 | c.805–825dup | M | 14 | Gigantism and hypopituitarism | GH | 766/− | MA |
| p.(Phe269–His275dup)† | |||||||
| 4 | c.910C>T | M | 20 | Gigantism and hypopituitarism | GH | 617/− | GA |
| p.(Arg304*)† |
The accession number used for the AIP transcript is NM_003977.2.
fs, frameshift; F, female; M, male; GH, growth hormone; PRL, prolactin; IGF-I, insulin-like growth factor I; MA, macro-adenoma (10 mm<diameter<40mm); GA, giant adenoma (≥40 mm diameter); AIPmut, AIP mutation; N/A, not available.
†Mutations encountered in FIPA patients.
AIP and AhR expression in the patients’ fibroblasts
As shown in Fig. 1B, AIP mRNA expression was significantly lower (∼50%) in M1 and M2 fibroblasts than in controls. AIP transcript levels in M3 and M4 fibroblasts were similar to control values. We examined whether nonsense-mediated mRNA decay (NMD) could explain this low mRNA level in M2 fibroblasts. NMD is a translation-coupled mechanism that eliminates mutated mRNAs that contain premature translation–termination codons located before the last exon. Indeed, as M2 resulted in a premature termination codon located in the third exon, NMD may account for the M2 AIP mRNA degradation and thus for the lack of truncated M2 protein (Silva et al. 2009). The presence of heterozygous AIP mutation in genomic DNA from M2 fibroblasts was confirmed, but direct cDNA sequencing failed to detect mutant transcripts (data not shown). Treatment of M2 fibroblasts with emetine, a potent NMD inhibitor, stabilized the mutant mRNA expressed from the defective allele and significantly increased the total amount of AIP mRNA, as measured by RT-qPCR (Supplementary Fig. S1).
AIP protein expression was lower (∼50%) in all four patients’ fibroblasts than in controls (Fig. 1C and D). No truncated or aberrant proteins were detected by WB with an antibody recognizing the N-terminal part of AIP. Interestingly, AIP protein expression was completely abolished in the pituitary adenoma of patient 1 (Fig. 1E). The subnormal AIP protein expression in patients’ fibroblasts, together with the lack of truncated forms, prompted us to examine the expression of AIP mutants in a translation-coupled-to-transcription assay. As expected, the M1 mutant could not be translated. We confirmed that the M3 and M4 mutants could be translated in vitro, into a longer and a shorter protein, respectively (Supplementary Fig. S2).
Analysis of AhR mRNA (Fig. 2A) and protein (Fig. 2B and C) showed that the expression levels in the patients’ fibroblasts were similar to control values, with variable protein expression levels in multiple WB experiments (Fig. 2C). ARNT transcript levels were also similar in AIP-mutated and control fibroblasts (data not shown). Thus, heterozygous AIP mutations do not alter intracellular AhR or ARNT levels in this cell model.
AhR expression in the fibroblasts of AIP-mutated patients and controls. (A) AhR transcript levels were similar in mutated and control fibroblasts, as determined by RT-qPCR. Results are expressed as the mean percentage of control ± s.e.m. of at least six independent determinations performed in duplicate. Values obtained with fibroblasts from the four controls were pooled and arbitrarily set at 100%. (B) AhR protein expression in WB experiments did not differ between AIP-mutated fibroblasts and controls. The positive control on this representative WB corresponds to protein extracts of HEK293T cells transfected or not transfected with a human AhR expression vector. (C) Quantitative analysis of six different WB experiments confirmed this result. Specific signals for AhR were normalized by the infrared fluorescence of α-tubulin. Data are presented as the AhR/α-tubulin ratio and as the percentage of control ± s.e.m. Values obtained with fibroblasts from the four controls were pooled and arbitrarily set at 100%.
Citation: Endocrine-Related Cancer 23, 5; 10.1530/ERC-16-0041
Expression of AhR target genes in AIP-mutated fibroblasts
To explore the impact of AIP mutations on AhR signaling, we studied the expression of CYP1B1, a prototypical AhR target gene that encodes a xenobiotic-metabolizing enzyme. As shown in Fig. 3A, B, C and D, fibroblasts exposure to the AhR ligand kynurenine resulted in a three- to four-fold increase in CYP1B1 transcript levels, as measured by RT-qPCR. Both 3′,4′-dimethoxyflavone (3,4-DMF), an AhR antagonist (Fig. 3A), and actinomycin D (ActD) (Fig. 3B) totally abolished the kynurenine-dependent increase in CYP1B1 transcript levels, indicating direct transcriptional control of CYP1B1 expression by AhR. This transcriptional response was time- and concentration-dependent (Fig. 3C and D), with maximal induction after 3–8h and at a kynurenine concentration of 25μM. Basal CYP1B1 transcript levels tended to be lower than control values in AIP-mutated fibroblasts, especially those from patients 2 and 4 (Fig. 3E). After kynurenine exposure, CYP1B1 transcript and protein levels were significantly lower in M2 and M4 fibroblasts than in controls, while both transcript and protein levels in M1 and M3 cells were similar to control values (Fig. 3E and F). Interestingly, the AIP mutations had different impacts on the expression of AhRR (aryl hydrocarbon receptor repressor), another AhR-regulated gene. Basal and kynurenine-induced AhRR mRNA levels were both higher in mutated fibroblasts than in controls (Fig. 4). Together, these results suggest that basal and ligand-induced transcriptional activation of AhR is modified in AIP-mutated fibroblasts, in a gene-specific manner.
CYP1B1 expression in fibroblasts from AIP-mutated patients and controls. (A) Fibroblast CYP1B1 expression was induced by kynurenine, as shown by RT-qPCR. 3′,4′-dimethoxyflavone, an AhR antagonist, abolished this effect of kynurenine. Cells were treated with 60 μM kynurenine, with or without 10 μM 3′,4′-dimethoxyflavone, for 8 h. (B) Actinomycin D also reversed the effect of kynurenine on fibroblast CYP1B1 mRNA expression, as shown by RT-qPCR. Cells were treated with 60 μM kynurenine, with or without 0.4 μM actinomycin D, for 8 h. (C) Kynurenine induced CYP1B1 expression in a time-dependent manner, peaking after 8 h of treatment. Cells were incubated for various times in the presence or absence of 60 μM kynurenine. (D) Kynurenine induced CYP1B1 expression in a concentration-dependent manner, peaking at 25 μM. Cells were incubated with various kynurenine concentrations for 8 h. (E) CYP1B1 transcript levels measured by RT-qPCR were lower in the fibroblasts of patients 2 and 4 than in controls, the difference reaching statistical significance only after exposure to 60 μM kynurenine for 8 h. (F) RT-qPCR results were confirmed by WB. Results of RT-qPCR are expressed as the mean percentage of control ± s.e.m. of at least four (A–D) or six (E) independent determinations performed in duplicate; values of vehicle-treated cells being arbitrarily set at 100%. *P<0.05, **P<0.01, ***P<0.001. Kynu, kynurenine; 3,4-DMF, 3′,4′-dimethoxyflavone; ActD, actinomycin D.
Citation: Endocrine-Related Cancer 23, 5; 10.1530/ERC-16-0041
AhRR expression in fibroblasts from AIP-mutated patients and controls. Basal and kynurenine-induced AhRR mRNA expression was significantly higher in AIP-mutated fibroblasts than in controls, as measured by RT-qPCR. Results are expressed as the mean percentage of control ± s.e.m. of at least six independent determinations performed in duplicate. Values obtained with fibroblasts from the four controls were pooled, and values for untreated cells were arbitrarily set at 100%. ****P<0.0001.
Citation: Endocrine-Related Cancer 23, 5; 10.1530/ERC-16-0041
Effect of AIP overexpression on Cyp1b1 mRNA expression in GH3 cells
As AIP mutations predispose humans to pituitary adenomas, we examined whether these mutations altered AhR signaling in the somatolactotroph GH3 cell line. As M2 AIP mRNA is degraded through an NMD mechanism, only the M1, M3, and M4 mutants were transfected into GH3 cells in order to examine AhR target gene expression. WB analysis confirmed that GH3 cells were successfully transfected (Fig. 5A). First, we used RT-qPCR to measure Cyp1b1 transcript levels after transfection of human wild-type (WT) and mutant AIP. Overexpression of neither WT nor mutant AIP significantly altered basal expression of the Cyp1b1 gene (Fig. 5B). Exposure to kynurenine induced a three-fold increase in Cyp1b1 mRNA expression. Cells transfected with WT AIP showed a larger increase in Cyp1b1 mRNA on kynurenine exposure than those transfected with either the empty vector (EV) or the mutant AIP vectors (Fig. 5B). Thus, overexpression of WT AIP, but not the AIP mutants, increased kynurenine-induced AhR target gene expression.
Cyp1b1 expression in GH3 cells after overexpression of wild-type and mutant AIP and knockdown of endogenous Aip. (A) WB analysis of AIP expression in GH3 cells transfected with WT AIP, mutant AIP, nontargeting siRNA (NT siRNA), or Aip siRNA. The WB shows overexpression of WT AIP, lack of the truncated AIP mutants, and successful knockdown of endogenous rat Aip by Aip siRNA. Protein extraction was performed 48 h post-transfection. (B) Kynurenine-induced Cyp1b1 expression in GH3 cells, as measured by RT-qPCR. Cells overexpressing wild-type AIP exhibited a large increase in Cyp1b1 transcript levels after kynurenine exposure than cells transfected with an empty vector (P<0.0001) or mutant AIP (P<0.0001 for WT AIP vs M1 and WT AIP vs M3 and P<0.001 for WT AIP vs M4). GH3 cells were transfected with either the empty pcDNATM3.1 vector or AIP-encoding vectors. After 1day, cells were deprived of tryptophan for 24 h and then exposed to 60 μM l-kynurenine for 3 h. RNA extraction was performed 48 h post-transfection. (C) Transfection of GH3 cells with Aip siRNA for 48 h resulted in successful knockdown of endogenous rat Aip, with a 70% decrease in Aip transcript levels in RT-qPCR experiments. (D) Knockdown of endogenous Aip attenuated the increase in Cyp1b1 transcript levels after kynurenine exposure. Cells were transiently transfected with Aip siRNA and nontargeting siRNA. After 48 h, cells were left either untreated or treated with 60 μM kynurenine for 3 h before RNA extraction. Changes in Cyp1b1 mRNA expression were detected by RT-qPCR. Results are expressed as the mean percentage of control ± s.e.m. of at least four (B) or six (C–D) independent determinations performed in duplicate, values for untreated cells transfected with an empty vector (B) or with nontargeting siRNA (C–D) being arbitrarily set at 100%. ***P<0.001, ****P<0.0001. EV, empty vector; WT, wild-type; NT, nontargeting.
Citation: Endocrine-Related Cancer 23, 5; 10.1530/ERC-16-0041
Effect of Aip knockdown on Cyp1b1 mRNA expression in GH3 cells
To further investigate the influence of the AIP expression level on AhR signaling, we measured Cyp1b1 transcript levels after inhibiting endogenous Aip gene expression with small interfering RNA (siRNA). As shown in Fig. 5A and C, a 70% decrease in endogenous AIP mRNA and a 60% decrease in endogenous AIP protein were observed, in both the absence and presence of kynurenine. Even partial silencing of Aip significantly reduced the increase in Cyp1b1 transcript levels after kynurenine exposure (Fig. 5D). Basal Cyp1b1 transcript levels were reduced by Aip knockdown but the difference did not reach statistical significance (P=0.05). Together, these results show that a reduction in AIP expression impairs AhR transcriptional activity in GH3 cells.
Impact of AIP on GH3 cells proliferation and GH secretion
We observed that in GH3 cells transfected with the empty vector, expressing only endogenous AIP protein, kynurenine treatment decreased cell proliferation (Supplementary Fig. S3). In untreated cells, transient transfection of WT AIP reduced cell proliferation, compared with the empty vector control. Kynurenine treatment reinforced this effect of WT AIP overexpression (Supplementary Fig. S3).
We further performed GH secretion assays in transfected GH3 cells after kynurenine stimulation. We observed a decrease in GH secretion in cells transfected with AIP mutants compared with cells transfected with either empty vector or WT AIP (Fig. 6A). A similar decrease in GH secretion was observed after Aip knockdown with siRNA in kynurenine-treated GH3 cells (Fig. 6B).
Effect of AIP mutants and Aip knockdown on kynurenine-induced GH secretion in GH3 cells. (A) GH secretion was significantly lower in kynurenine-treated GH3 cells transfected with M1, M3, and M4 mutants. Cells were transfected with either the empty pcDNATM3.1 vector or AIP-encoding vectors, deprived of tryptophan and exposed to 60 μM l-kynurenine for 48 h. GH concentrations were measured in the medium 48 h post-transfection. Results are expressed as the mean percentage of control ± s.e.m., values for untreated cells transfected with an empty vector being arbitrarily set at 100%. (B) Aip knockdown in GH3 cells induced a decrease in GH secretion after a 48 h exposure to kynurenine. Cells were transiently transfected with Aip siRNA and nontargeting siRNA. After 48 h, cells were left either untreated or treated with 60 μM kynurenine for 48 h before GH secretion analysis. Results are expressed as the mean percentage of control ± s.e.m., values for cells transfected with nontargeting siRNA being arbitrarily set at 100%. *P<0.05, **P<0.01, ***P<0.001. EV, empty vector; WT, wild-type; NT, nontargeting.
Citation: Endocrine-Related Cancer 23, 5; 10.1530/ERC-16-0041
Discussion
We describe for the first time the functional impact of human AIP mutations on AhR signaling, using fibroblasts from patients with heterozygous AIP mutations. Previously, the role of AIP in AhR signaling had only been analyzed in murine cell lines overexpressing exogenous AIP, and none of the AIP mutations previously identified in patients with pituitary adenomas had been examined for their impact on the AhR pathway. The AIP mutations studied here have already been reported (Vierimaa et al. 2006, Personnier et al. 2011, Cazabat et al. 2012) and are considered pathogenic (Beckers et al. 2013, Hernández-Ramírez et al. 2015a). Two of these mutations (M3 and M4) have been reported in patients with familial isolated pituitary adenomas. Interestingly, the four mutations studied here had different predicted impacts on AIP protein expression, two of them (M2 and M4) potentially leading to a truncated protein lacking the C-terminal domain required for AIP–AhR interaction.
We found that the heterozygous AIP mutations resulted in decreased AIP protein expression, by comparison with AIP wild-type fibroblasts. In the case of mutations M1 and M2, this is easily explained by the loss of the translation initiation codon and by an NMD mechanism, respectively. Regarding mutations M3 and M4, the reduced AIP expression and the nonexpression of a truncated AIP protein could be linked to a shorter protein half-life due to enhanced degradation (Hernández-Ramírez LC et al. 2015b). The absence of AIP protein in the pituitary adenoma of patient 1 (M1) points to a ‘second hit’ (Knudson’s hypothesis (Knudson 1971)), which could be a loss of heterozygosity, as observed elsewhere in several AIP-mutated tumors (Lecoq et al. 2015).
Interestingly, this decrease in endogenous AIP protein expression did not affect the AhR expression level. As AIP has been shown to protect the ligand-free form of murine AhR against ubiquitination (Meyer et al. 1999, Kazlauskas et al. 2000, LaPres et al. 2000), a decrease in AhR protein levels might have been expected in AIP-mutated fibroblasts. This unexpected finding might be explained by different roles of AIP in human vs mouse AhR complexes (Ramadoss et al. 2004). In addition, other chaperone proteins important for AhR stabilization, such as HSP90, might compensate for the AIP deficiency (Chen et al. 1997).
To explore the impact of AIP mutations on AhR signaling, we studied the expression of CYP1B1, a prototypical AhR target gene, before and after kynurenine exposure. Kynurenine was chosen to stimulate AhR because this oncometabolite, being an endogenous AhR ligand, is likely to be devoid of the toxic effects of TCDD and has been shown to activate AhR in tumor cells (Opitz et al. 2011). We confirm that kynurenine also activates AhR-dependent CYP1B1 transcription in patients’ fibroblasts and in pituitary cells, at concentrations as low as 25μM. However, we have chosen the 60μM concentration which corresponds to the concentration measured in the media of human glioma cells and which was able to stimulate the AhR pathway in these cells (Opitz et al. 2011). Interestingly, we have found that the tryptophan-metabolizing enzyme IDO2 required for kynurenine synthesis is also expressed in somatotroph adenomas (not shown). Thus, locally produced kynurenine could play a role in AIP-related pituitary tumorigenesis but its pituitary production remains to be demonstrated.
We show for the first time that heterozygous human AIP mutations modify both basal and ligand-induced AhR transcriptional activity in human fibroblasts. This effect is particularly observed after kynurenine exposure for CYP1B1, which is similar to what is reported in other studies assessing the in vitro transcriptional activity of AhR (Ma et al. 1997, Pollenz et al. 2005) and with other transcription factors such as nuclear receptors (Laenger et al. 2009). The differential impact of AIP mutations on CYP1B1 transcription is intriguing. The exact molecular mechanisms by which these mutations alter the transcriptional regulation of CYP1B1 by AhR is currently unknown. In the M2 and M4 cell lines, AIP mutations lead in theory to truncation of the C-terminal part of the AIP protein required for interaction with AhR. The resulting mutant proteins could act as dominant negatives of the WT AIP protein by interfering with the AhR-dependent transcription. Nevertheless, the existence of these truncated proteins is questionable since we were unable to detect them by Western blot analysis, but this does not formally rule out their transient expression. Alternatively, the M2 and M4 mutations may simply result in the absence of mutated AIP protein and thus lead to AIP haploinsufficiency, which could be responsible for the observed impact on AhR target gene transcription. Previous studies have shown that AIP translocates to the nucleus together with AhR in human cells (Ramadoss et al. 2004), and that variations in AIP expression may enhance or repress AhR transcriptional activity (Ma et al. 1997, Carver et al. 1998, Hollingshead et al. 2004). Thus, AIP deficiency might modify AhR nuclear translocation and/or AhR binding to ARNT and coregulators. We found that different AIP mutations had different impacts on CYP1B1 and AhRR transcription, thus underlining the complexity of AhR signaling; indeed, this transcription factor can bind several response elements and recruit various coactivators/repressors in the regulatory sequences of its target genes (Hankinson 2005).
Several lines of evidence suggest a functional role of the AhR signaling pathway in the somatotroph cells. First, using the well-characterized somatolactotroph GH3 cells, we found that the overexpression of WT AIP, but not the AIP mutants, enhanced AhR transcriptional activity, while knockdown of endogenous Aip gave the opposite results. This suggests that AIP deficiency in these pituitary cells may lead to downregulation of AhR transcriptional activity. Secondly, activation of AhR signaling by kynurenine had antiproliferative action in this cell model. Finally, unlike cAMP pathway activation by forskolin (Formosa et al. 2013), AhR activation by kynurenine did not modify basal GH secretion from untransfected GH3 cells expressing normal amounts of endogenous AIP (not shown). However, both AIP mutants and Aip knockdown decreased GH secretion from kynurenine-treated GH3 cells. A similar observation has been previously reported after Aip knockdown in GH3 cells exposed to forskolin and IBMX (Formosa et al. 2013). Thus, AIP haploinsufficiency, both after cAMP pathway and AhR pathway activation, decreases GH secretion from GH3 cells. These results seem paradoxical considering the role of AIP mutations in somatotroph tumorigenesis. Nevertheless, similarly to what has been recently reported for ARMC5 mutations (Assié et al. 2013), we can hypothesize that the inactivation of AIP might be, per se, associated with a partial dedifferentiation of pituitary somatotroph cells leading to altered capacity of mutated cells to produce GH. The increased proliferation rate of somatotroph cells bearing AIP mutations (Leontiou et al. 2008) leads to tumor formation and in fine to GH excess.
In conclusion, in patients’ fibroblasts bearing endogenous heterozygous AIP mutations and in transfected pituitary GH3 cells, AIP mutations affect AhR transcriptional activity in a gene- and tissue-dependent manner. How altered AhR signaling contributes to pituitary tumorigenesis remains to be further investigated.
Supplementary data
This is linked to the online version of the paper at http://dx.doi.org/10.1530/ERC-16-0041.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this review.
Funding
This work was supported by grants from Inserm, Université Paris-Sud, ITMO Cancer-Plan Cancer 2009–2013, and Fondation ARC. A L L was the recipient of a fellowship from the French Endocrine Society (Société Française d’Endocrinologie) and P K was the recipient of a Contrat d’Interface Inserm 2014.
Author contribution statement
A L L, M L, P C and P K contributed to the conception and design of the research; A L L, S V, M H and P Z performed the experiments; J B performed AIP sequencing; A B obtained fibroblasts from the patients’ skin biopsies; J Y, P C and P K referred the patients; A L L, S V, M H and P K analyzed the data; A L L, S V, M L, P C and P K interpreted the results of the experiments; A L L, S V and P K prepared the figures; A L L and P K drafted the manuscript; A L L, S V, M L, P C and P K edited and revised the manuscript; A L L, S V, J B, J Y, M L, P C and P K approved the final version of the manuscript.
Acknowledgments
We thank Marie Legendre for providing the GH3 cells used in this study, Benoît Lambert and Jean-François Papon for their help in the search for healthy controls, Clovis Adam for providing tissue samples, and Elodie Dupuis for excellent technical assistance with AIP sequencing.
References
Assié G , Libé R , Espiard S , Rizk-Rabin M , Guimier A , Luscap W , Barreau O , Lefèvre L , Sibony M & Guignat L et al. 2013 ARMC5 mutations in macronodular adrenal hyperplasia with Cushing’s syndrome. New England Journal of Medicine 369 2105–2114. (doi:10.1056/NEJMoa1304603)
Barouki R , Coumoul X & Fernandez-Salguero PM 2007 The aryl hydrocarbon receptor, more than a xenobiotic-interacting protein. FEBS Letters 581 3608–3615. (doi:10.1016/j.febslet.2007.03.046)
Beckers A , Aaltonen LA , Daly AF & Karhu A 2013 Familial isolated pituitary adenomas (FIPA) and the pituitary adenoma predisposition due to mutations in the aryl hydrocarbon receptor interacting protein (AIP) gene. Endocrine Reviews 34 239–277. (doi:10.1210/er.2012-1013)
Bell DR & Poland A 2000 Binding of aryl hydrocarbon receptor (AhR) to AhR-interacting protein. The role of hsp90. Journal of Biological Chemistry 275 36407–36414. (doi:10.1074/jbc.M004236200)
Cannavò S , Ferraù F , Ragonese M , Curtò L , Torre ML , Magistri M , Marchese A , Alibrandi A & Trimarchi F 2010 Increased prevalence of acromegaly in a highly polluted area. European Journal of Endocrinology 163 509–513. (doi:10.1530/EJE-10-0465)
Carver LA , LaPres JJ , Jain S , Dunham EE & Bradfield CA 1998 Characterization of the Ah receptor-associated protein, ARA9. Journal of Biological Chemistry 273 33580–33587. (doi:10.1074/jbc.273.50.33580)
Cazabat L , Bouligand J , Salenave S , Bernier M , Gaillard S , Parker F , Young J , Guiochon-Mantel A & Chanson P 2012 Germline AIP mutations in apparently sporadic pituitary adenomas: prevalence in a prospective single-center cohort of 443 patients. Journal of Clinical Endocrinology and Metabolism 97 E663–E670. (doi:10.1210/jc.2011-2291)
Chen HS , Singh SS & Perdew GH 1997 The Ah receptor is a sensitive target of geldanamycin-induced protein turnover. Archives of Biochemistry and Biophysics 348 190–198. (doi:10.1006/abbi.1997.0398)
Formosa R , Xuereb-Anastasi A & Vassallo J 2013 Aip regulates cAMP signalling and GH secretion in GH3 cells. Endocrine-Related Cancer 20 495–505. (doi:10.1530/ERC-13-0043)
Hankinson O 2005 Role of coactivators in transcriptional activation by the aryl hydrocarbon receptor. Archives of Biochemistry and Biophysics 433 379–386. (doi:10.1016/j.abb.2004.09.031)
Heliövaara E , Raitila A , Launonen V , Paetau A , Arola J , Lehtonen H , Sane T , Weil RJ , Vierimaa O & Salmela P et al. 2009 The expression of AIP-related molecules in elucidation of cellular pathways in pituitary adenomas. American Journal of Pathology 175 2501–2507. (doi:10.2353/ajpath.2009.081131)
Hernández-Ramírez LC , Gabrovska P , Dénes J , Stals K , Trivellin G , Tilley D , Ferraù F , Evanson J , Ellard S & Grossman AB et al. 2015 aLandscape of familial isolated and young-onset pituitary adenomas: prospective diagnosis in AIP mutation carriers. Journal of Clinical Endocrinology and Metabolism 100 E1242–E1254. (doi:10.1210/jc.2015-1869)
Hernández-Ramírez LC , Martucci F , Ferraù F , Morgan Rhodri ML , Trivellin G , Begum F , Tilley D , Ramos-Guajardo N , Iacovazzo D & Prodromou C et al. 2015 b In The Enhanced Proteosomal Degradation of AIP Mutant Proteins Is a Mechanism for AIP Deficiency in AIP mutation-Associated Pituitary Adenomas. San Diego, CA, USA.
Hollingshead BD , Petrulis JR & Perdew GH 2004 The aryl hydrocarbon (Ah) receptor transcriptional regulator hepatitis B virus X-associated protein 2 antagonizes p23 binding to Ah receptor-Hsp90 complexes and is dispensable for receptor function. Journal of Biological Chemistry 279 45652–45661. (doi:10.1074/jbc.M407840200)
Jaffrain-Rea M-L , Angelini M , Gargano D , Tichomirowa MA , Daly AF , Vanbellinghen J-F , D’Innocenzo E , Barlier A , Giangaspero F & Esposito V et al. 2009 Expression of aryl hydrocarbon receptor (AHR) and AHR-interacting protein in pituitary adenomas: pathological and clinical implications. Endocrine-Related Cancer 16 1029–1043. (doi:10.1677/ERC-09-0094)
Kazlauskas A , Poellinger L & Pongratz I 2000 The immunophilin-like protein XAP2 regulates ubiquitination and subcellular localization of the dioxin receptor. Journal of Biological Chemistry 275 41317–41324. (doi:10.1074/jbc.M007765200)
Knudson AG 1971 Mutation and cancer: statistical study of retinoblastoma. PNAS 68 820–823. (doi:10.1073/pnas.68.4.820)
Laenger A , Lang-Rollin I , Kozany C , Zschocke J , Zimmermann N , Rüegg J , Holsboer F , Hausch F & Rein T 2009 XAP2 inhibits glucocorticoid receptor activity in mammalian cells. FEBS Letters 583 1493–1498. (doi:10.1016/j.febslet.2009.03.072)
LaPres JJ , Glover E , Dunham EE , Bunger MK & Bradfield CA 2000 ARA9 modifies agonist signaling through an increase in cytosolic aryl hydrocarbon receptor. Journal of Biological Chemistry 275 6153–6159. (doi:10.1074/jbc.275.9.6153)
Le Billan F , Khan JA , Lamribet K , Viengchareun S , Bouligand J , Fagart J & Lombès M 2015 Cistrome of the aldosterone-activated mineralocorticoid receptor in human renal cells. FASEB Journal 29 3977–3989. (doi:10.1096/fj.15-274266)
Lecoq A-L , Kamenický P , Guiochon-Mantel A & Chanson P 2015 Genetic mutations in sporadic pituitary adenomas – what to screen for? Nature Reviews Endocrinology 11 43–54. (doi:10.1038/nrendo.2014.181)
Leontiou CA , Gueorguiev M , van der Spuy J , Quinton R , Lolli F , Hassan S , Chahal HS , Igreja SC , Jordan S & Rowe J et al. 2008 The role of the aryl hydrocarbon receptor-interacting protein gene in familial and sporadic pituitary adenomas. Journal of Clinical Endocrinology and Metabolism 93 2390–2401. (doi:10.1210/jc.2007-2611)
Ma Q & Whitlock JP Jr 1997 A novel cytoplasmic protein that interacts with the Ah receptor, contains tetratricopeptide repeat motifs, and augments the transcriptional response to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Journal of Biological Chemistry 272 8878–8884.
Meyer BK & Perdew GH 1999 Characterization of the AhR-hsp90-XAP2 core complex and the role of the immunophilin-related protein XAP2 in AhR stabilization. Biochemistry 38 8907–8917. (doi:10.1021/bi982223w)
Morgan RML , Hernández-Ramírez LC , Trivellin G , Zhou L , Roe SM , Korbonits M & Prodromou C 2012 Structure of the TPR domain of AIP: lack of client protein interaction with the C-terminal α-7 helix of the TPR domain of AIP is sufficient for pituitary adenoma predisposition. PLoS ONE 7 e53339. (doi:10.1371/journal.pone.0053339)
Murray IA , Patterson AD & Perdew GH 2014 Aryl hydrocarbon receptor ligands in cancer: friend and foe. Nature Reviews Cancer 14 801–814. (doi:10.1038/nrc3846)
Nguyen LP & Bradfield CA 2008 The search for endogenous activators of the aryl hydrocarbon receptor. Chemical Research in Toxicology 21 102–116. (doi:10.1021/tx7001965)
Opitz CA , Litzenburger UM , Sahm F , Ott M , Tritschler I , Trump S , Schumacher T , Jestaedt L , Schrenk D & Weller M et al. 2011 An endogenous tumour-promoting ligand of the human aryl hydrocarbon receptor. Nature 478 197–203. (doi:10.1038/nature10491)
Personnier C , Cazabat L , Bertherat J , Gaillard S , Souberbielle J-C , Habrand JL , Dufour C , Clauser E , SainteRose C & Polak M 2011 Clinical features and treatment of pediatric somatotropinoma: case study of an aggressive tumor due to a new AIP mutation and extensive literature review. Hormone Research in Pædiatrics 75 392–402. (doi:10.1159/000327831)
Pesatori AC , Baccarelli A , Consonni D , Lania A , Beck-Peccoz P , Bertazzi PA & Spada A 2008 Aryl hydrocarbon receptor-interacting protein and pituitary adenomas: a population-based study on subjects exposed to dioxin after the Seveso, Italy, accident. European Journal of Endocrinology 159 699–703. (doi:10.1530/EJE-08-0593)
Pollenz RS & Dougherty EJ 2005 Redefining the role of the endogenous XAP2 and C-terminal hsp70-interacting protein on the endogenous Ah receptors expressed in mouse and rat cell lines. Journal of Biological Chemistry 280 33346–33356. (doi:10.1074/jbc.M506619200)
Puga A , Ma C & Marlowe JL 2009 The aryl hydrocarbon receptor cross-talks with multiple signal transduction pathways. Biochemical Pharmacology 77 713–722. (doi:10.1016/j.bcp.2008.08.031)
Raitila A , Lehtonen HJ , Arola J , Heliövaara E , Ahlsten M , Georgitsi M , Jalanko A , Paetau A , Aaltonen LA & Karhu A 2010 Mice with inactivation of aryl hydrocarbon receptor-interacting protein (Aip) display complete penetrance of pituitary adenomas with aberrant ARNT expression. American Journal of Pathology 177 1969–1976. (doi:10.2353/ajpath.2010.100138)
Ramadoss P , Petrulis JR , Hollingshead BD , Kusnadi A & Perdew GH 2004 Divergent roles of hepatitis B virus X-associated protein 2 (XAP2) in human versus mouse Ah receptor complexes. Biochemistry 43 700–709. (doi:10.1021/bbib35827v)
Schmidt JV & Bradfield CA 1996 Ah receptor signaling pathways. Annual Review of Cell and Developmental Biology 12 55–89. (doi:10.1146/annurev.cellbio.12.1.55)
Silva AL & Romão L 2009 The mammalian nonsense-mediated mRNA decay pathway: to decay or not to decay! Which players make the decision? FEBS Letters 583 499–505. (doi:10.1016/j.febslet.2008.12.058)
Steyn FJ , Huang L , Ngo ST , Leong JW , Tan HY , Xie TY , Parlow AF , Veldhuis JD , Waters MJ & Chen C 2011 Development of a method for the determination of pulsatile growth hormone secretion in mice. Endocrinology 152 3165–3171. (doi:10.1210/en.2011-0253)
Trivellin G & Korbonits M 2011 AIP and its interacting partners. Journal of Endocrinology 210 137–155. (doi:10.1530/JOE-11-0054)
Tuominen I , Heliövaara E , Raitila A , Rautiainen M-R , Mehine M , Katainen R , Donner I , Aittomäki V , Lehtonen HJ & Ahlsten M et al. 2014 AIP inactivation leads to pituitary tumorigenesis through defective Gαi-cAMP signaling. Oncogene. (doi:10.1038/onc.2014.50)
Vierimaa O , Georgitsi M , Lehtonen R , Vahteristo P , Kokko A , Raitila A , Tuppurainen K , Ebeling TML , Salmela PI & Paschke R et al. 2006 Pituitary adenoma predisposition caused by germline mutations in the AIP gene. Science 312 1228–1230. (doi:10.1126/science.1126100)
