Abstract
Germline aryl hydrocarbon receptor interacting protein (AIP) gene mutations confer a predisposition to pituitary adenoma (PA), predominantly GH-secreting (GH-PA). As recent data suggest a role for AIP in the pathogenesis of sporadic GH-PA and their response to somatostatin analogues (SSA), the expression of AIP and its partner, aryl hydrocarbon receptor (AHR), was determined by semiquantitative immunohistochemistry scoring in 62 sporadic GH-PA (37 treated with SSA preoperatively). The influence of Gsp status was studied in a subset of tumours (n=39, 14 Gsp+) and six GH-PA were available for primary cultures. AIP and AHR were detected in most cases, with a positive correlation between AIP and cytoplasmic AHR (P=0.012). Low AIP expression was significantly more frequent in untreated vs SSA-treated tumours (44.0 vs 20.5%, P=0.016). AHR expression or localisation did not differ between the two groups. Similarly, in vitro octreotide induced a median twofold increase in AIP expression (range 1.2–13.9, P=0.027) in GH-PA. In SSA-treated tumours, the AIP score was significantly higher in the presence of preoperative IGF1 decrease or tumour shrinkage (P=0.008 and P=0.014 respectively). In untreated tumours, low AIP expression was significantly associated with invasiveness (P=0.028) and suprasellar extension (P=0.019). The only effect of Gsp status was a significantly lower nuclear AHR score in Gsp+ vs Gsp− tumours (P=0.025), irrespective of SSA. In conclusion, AIP is involved in the aggressiveness of sporadic GH-PA, regardless of Gsp status, and AIP up-regulation in SSA-treated tumours is associated with a better preoperative response, with no clear role for AHR.
Introduction
The aryl hydrocarbon receptor interacting protein (AIP) gene has been identified as a predisposing gene for the development of pituitary adenomas (PA) (Vierimaa et al. 2006). AIP is a pituitary tumour suppressor gene, with germline inactivating mutations being associated with somatic loss of heterozygosity in the corresponding tumours. A particular preponderance of growth hormone (GH)-secreting PA (GH-PA) in the setting of germline AIP mutations (AIPmut) has been reported by clinical studies worldwide (Beckers et al. 2013). In our experience, germline AIPmut accounts for 50% of familial isolated PA kindreds with homogeneous somatotropinomas (Daly et al. 2007) and 80% of AIPmut patients with PA have GH-PA (Daly et al. 2010). AIPmut GH-PA can also present as an apparently sporadic acromegaly or gigantism in young patients, especially in a paediatric context (Beckers et al. 2013). When compared with their non-AIPmut counterparts, AIPmut somatotropinomas are commonly more aggressive and more frequently resistant to somatostatin analogues (SSA; Daly et al. 2010).
There is a growing body of evidence that AIP down-regulation may contribute to the pathogenesis of sporadic PA, regardless of AIPmut status. Although somatic AIP mutations have not been reported to date (Barlier et al. 2007, Iwata et al. 2007, Jaffrain-Rea et al. 2009), AIP expression is frequently reduced in invasive GH-PA (Jaffrain-Rea et al. 2009), and loss of AIP immunostaining has been proposed as a marker of tumour aggressiveness in sporadic cases (Kasuki Jomori de Pinho et al. 2011). There is also recent evidence that AIP expression is increased by SSA (Jaffrain-Rea et al. 2010, Chahal et al. 2012) and may predict the post-operative response to SSA in acromegalic patients (Kasuki et al. 2012). This suggests that loss of AIP function or expression may contribute to pharmacological resistance in GH-PA.
We have previously observed that AIP mutations or down-regulation in PA is frequently accompanied by a reduced expression of its best characterised partner, the aryl hydrocarbon receptor (AHR; Jaffrain-Rea et al. 2009). AHR is stabilised in the cytoplasm into a latent AIP/AHR/heat-shock protein 90/p23 complex (Petrulis & Perdrew 2002). Upon activation by exogenous ligands, it translocates to the nucleus and exerts transcriptional effects after heterodimerisation with the aryl hydrocarbon receptor nuclear translocator (ARNT; Beischlag et al. 2008). The AHR/ARNT complex mediates the detoxifying effects of AHR and is involved in endocrine disruption (Beischlag et al. 2008). Endogenous functions of AHR include the control of apoptosis and cell cycle proliferation (Nguyen & Bradfield 2008) and AHR signalling may up-regulate p27Kip1 (Marlowe & Puga 2005), which is also increased by SSA in GH-PA (Ferrante et al. 2006, Hubina et al. 2006). However, the potential effects of SSA on AHR expression and localisation in GH-PA is unknown.
The best understood genetic event in sporadic GH-PA is the presence of somatic activating mutations of the GNAS1 (GNAS) gene – the Gsp oncogene – which induce a constitutive activity of the cAMP pathway, an important target of SSA (Lania et al. 2003). As cAMP is an endogenous activator of AHR (Oesch-Bartlomowicz et al. 2005) and cAMP signalling can be modulated by AIP through direct interactions with specific phosphodiesterases (PDEs; Bolger et al. 2003, de Oliveira et al. 2007), Gsp may also theoretically affect AIP and/or AHR expression.
We therefore aimed to further evaluate the expression of AIP and AHR in a large series of sporadic GH-PA, taking into account the presence of preoperative SSA treatment and the potential effects of Gsp mutations.
Subjects and methods
Patients and samples
A series of surgical samples from 62 GH-PA operated on in patients with sporadic acromegaly was retrospectively studied, including 52 cases from three European centres (Neuromed, Pozzilli and University/Hospital of Padova, Italy; CHU, University of Liège, Belgium) and ten tumours with Gsp mutations (Gsp+) provided by additional two centres (University of Aix-Marseilles, France and University of Milan, Italy). Most tumours were collected between 2006 and 2011; archive material collected from 1997 to 2005 was included in order to ensure an appropriate ratio of SSA-treated and -untreated tumours and a sufficient number of Gsp+ samples. The study was approved by local ethics committees. Patients diagnosed before the age of 18 years, with a familial history of acromegaly or with germline AIP mutations, were excluded from the study, as well as those who received preoperative therapy with dopamine agonists only. Of the patients, 25 were males and 37 were females, with a mean age of 44.8±12.8 years at surgery (median 44.5 years and range 18–78 years). Patients' clinical, biological and neuroradiological data were recorded. GH, prolactin (PRL) and insulin-like growth factor 1 (IGF1) measurements were obtained for each centre. IGF1 levels were adjusted for age and expressed as percentage of upper normal limit (%ULN). Hormone data were incomplete in 12 cases. Most patients had macroadenomas, as defined by maximal tumour diameter exceeding 10 mm (52 out of 62 cases, 83.9%), a suprasellar extension was recorded in 29 out of 59 cases (49.1%) – not recorded in three cases – and 31 out of 62 cases (50%) were invasive. A total of 37 patients (59.7%) received SSA treatment before surgery (octreotide–LAR 20–40 mg monthly, n=21 or lanreotide 60–120 mg monthly, n=16), including four patients who also received dopamine agonists for mixed GH/PRL-secreting tumours. Median preoperative treatment duration was 6 months (range 3–108). The pharmacological response was assessed in terms of plasma GH and IGF1 reduction, hormone values at diagnosis and before surgery being available for 33 treated patients to calculate ΔGH and ΔIGF1 as percentages of hormone decrease. Treated patients were then divided into three groups on the basis of preoperative age-corrected IGF1 values (%ULN): group I, controlled (IGF1 normalisation for age, n=10); group II, partially controlled (non-normalised IGF1 with ΔIGF1 ≥30% when compared with pretreatment values, n=13) and group III, uncontrolled (ΔIGF1 <30% or IGF1 increase when compared with pretreatment values, n=10). Data on tumour shrinkage were also available for 26 patients (including one patient with no available ΔIGF1). Immunohistochemistry (IHC) for pituitary hormones was performed in each centre. Cell proliferation was evaluated in 54 cases by Ki67 immunostaining with the monoclonal MIB1 antibody, as described previously (Jaffrain-Rea et al. 2002).
Molecular and genetic analysis
Patients gave informed consent for genetic analysis. Direct AIP sequencing was performed on leukocyte DNA as previously reported (Daly et al. 2007, Occhi et al. 2010). A search for activating mutations of the GNAS1 gene at codons 201 and 227 (Gsp) was performed on tumour DNA at each centre, as described previously (Barlier et al. 1998, Lania et al. 1998, Occhi et al. 2011), primers and conditions for GNAS1 sequencing in the leading centre being available on demand. In a minority of patients lost to follow-up, AIP and GNAS1 sequencing was performed on tumour DNA obtained from paraffin-embedded material (QIAmp DNA FFPE tissue kit, Qiagen). Overall, AIP and GNAS1 sequences could be obtained in 40 and 39 cases respectively, 29 being characterised for both genes. No case had an AIP mutation, and no significant difference was retrospectively found between tumours characterised or not for AIP sequencing in terms of patient's age, gender, tumour volume, invasiveness and preoperative SSA treatment (data not shown). Gsp mutations were present in 14/39 cases (Gsp+).
Immunohistochemistry
IHC was performed at the University of L'Aquila as described previously (Jaffrain-Rea et al. 2009) using a mouse monoclonal anti-AIP antibody at a 1:500 dilution in all cases (clone 35-2, Novus Biologicals LLC, Littleton, CO, USA, distributed by DBA Italia, Milan, Italy) and a polyclonal rabbit anti-AHR antibody at a 1:50 dilution in 53 cases (sc-5579, Santa Cruz Biotechnology, distributed by DBA Italia). Negative controls were performed omitting the primary antibody and normal pituitary (NP) tissue adjacent to PA samples was used as positive controls. Immunostaining for AIP was classified semiquantitatively according to intensity (negative, 0; weak, 1; moderate, 2 and strong, 3) and expression pattern (patchy, 1 and diffuse, 2), with a final score being obtained by multiplying intensity×pattern (range 0–6) (Leontiou et al. 2008, Kasuki Jomori de Pinho et al. 2011). Low AIP immunostaining (low AIP–IHC) was defined by a semiquantitative score ≤2. AHR immunostaining was also classified semiquantitatively (negative, 0; weak, 1; moderate, 2 and strong, 3) in terms of cytoplasmic AHR (AHRc) and nuclear AHR (AHRn) localisation. Total AHR (AHRt) score was calculated by adding AHRc and AHRn scores. High AHRc and AHRn scores were defined as ≥2 and high AHR content (high AHR–IHC) by AHRt score ≥4. Photographs of slides were taken using a Zeiss Axioplan 2 microscope (Carl Zeiss Microimaging, Inc., USA) and a Leica DFC 320 digital camera (Leica GmBH, Germany).
In vitro study of GH-secreting PA
The study was previously approved by the local ethics committees at the Universities of Milan and Marseilles and informed consent was obtained from the patients. Fresh tissue was obtained by the transsphenoidal route from six untreated GH-PA and enzymatically dissociated in DMEM containing 2 mg/ml collagenase at 37 °C for 2 h, as described previously (Lania et al. 2004). Cells from two GH-PA were incubated with or without octreotide (10 nM) for 6, 24 and 48 h and lysed in the presence of protease inhibitors. On the basis of these experiments and due to limited amounts of cells, four additional tumours were studied at baseline and after 24 h of treatment. For each sample, 20 μg proteins were separated on 12% SDS–polyacrylamide gels and transferred to a nitrocellulose filter. Western blotting experiments were performed with the antibodies used for IHC reported hitherto, using a 1:1000 dilution for the anti-AIP antibody and a 1:500 dilution for the anti-AHR antibody respectively and HRP-linked secondary antibodies for protein detection. GAPDH was used as a housekeeping control. The resulting bands were evaluated with the image analysis program NIH Image J (NIH, Bethesda, MD, USA; http://rsbweb.nih.gov/ij).
Statistical analysis
Data are expressed in median (range) and statistical analyses were performed using Statview 5.01 Software for PC (SAS Institute, Cary, NC, USA). Data were analysed by non-parametric analysis, using the Mann–Whitney U and Kruskal–Wallis tests for 2 and ≥3 groups' comparisons respectively. The non-parametric Spearman's test was used to correlate AIP and AHR immunoscores. Results are given by ex-aequo corrections for the non-parametric analysis of immunoscores. Distribution of nominal values was compared by the χ2 test. Multiple logistic regression was performed to evaluate the influence of preoperative SSA where appropriate. In primary cell culture experiments, the Wilcoxon's rank test was used to compare AIP:GADPH and AHR:GADPH ratios in treated and untreated control cells. P<0.05 was considered significant.
Results
Study population
In order to look for potential bias in the selection of patients treated by with SSA before surgery, the characteristics of treated and untreated patients and tumours were retrospectively compared (Table 1). Female patients predominated in both groups. Although treated patients tended to be older than untreated patients, the difference was not significant (P=0.055). SSA-treated and -untreated groups had similar proportions of subjects with proven normal AIP sequencing (including young patients). Plasma GH, PRL and IGF1 (%ULN) levels at diagnosis were similar in both groups. No significant difference existed between the two groups in terms of tumour volume, invasiveness or functional phenotypes. Gsp+ tumours were also similarly distributed among treated and untreated patients. The significantly lower Ki67 index observed in treated vs untreated tumours (P=0.002) was ascribed to the anti-proliferative effect of SSA on somatotropinomas.
Comparison of clinical and genetic characteristics in 62 patients affected by sporadic acromegaly according to the presence or the absence of preoperative treatment with somatostatin analogues
| Untreated | Treated | P | |
|---|---|---|---|
| Patients (n) | 25 | 37 | |
| Gender | 9 M/16 F | 16 M/21 F | 0.568 |
| Age (years) | 42 (18–64) | 49 (21–78) | 0.055 |
| Macroadenomas | 21/25 (84.0%) | 31/37 (83.8%) | 0.982 |
| Suprasellar extension | 13/23 (56.5%) | 16/36 (44.4%) | 0.366 |
| Maximal tumour diameter (mm) | 13.0 (8.0–40.0) (18) | 15.0 (8.0–40.0) (31) | 0.901 |
| Invasive tumours | 12/25 (48.0%) | 19/37 (51.3%) | 0.796 |
| GH at diagnosis (ng/ml) | 19.6 (4.0–82.0) (23) | 22.4 (4.0–107.0) (33) | 0.120 |
| IGF1 at diagnosis (%ULN) | 203.6 (51.3–814.0) (21) | 255.1 (127.0–636.0) (32) | 0.155 |
| PRL at diagnosis (ng/ml) | 21.5 (12.7–92.5) (21) | 15.0 (3.2–253.2) (32) | 0.473 |
| Pituitary hormones IHC | 0.326 | ||
| GH | 14/25 | 21/37 | – |
| GH/PRL | 6/25 | 13/37 | – |
| GH/glycoP | 5/25 | 3/37 | – |
| Ki67 (%) | 1.3 (0.0–8.0) (21) | 0.8 (0.0–3.5) (33) | 0.002 |
| AIP genetics | 18/25 (72.0%) | 22/37 (59.4%) | 0.311 |
| Age <30 years old | 4/6 (66.6%) | 3/4 (75.0%) | – |
| Gsp+ | 5/16 (31.2%) | 9/23 (39.1%) | 0.614 |
Owing to the retrospective character of the study, some data were missing. For continuous variables with missing values, the number of available cases is indicated within parentheses for each item.
General immunohistochemical findings
Examples of AIP and AHR immunostaining in four representative cases and in a NP control are shown in Fig. 1. Some degree of AIP immunostaining was observed in all but two cases (96.8%). Diffuse cytoplasmic immunostaining was the most frequent pattern and low AIP expression was observed in 17 out of 62 cases (27.4%). Preoperative SSA was found to significantly influence AIP expression. Low AIP–IHC was twice as frequent in untreated vs SSA-treated GH-PA (44.0 vs 20.5%, P=0.016). A similar trend was seen taking only those with proven normal AIP status, although the difference did not reach significance (44.4 vs 22.7%, P=0.140). Hence, low AIP-expressing GH-PA were characterised in the whole series and in untreated cases taken as a subgroup (Table 2).
AIP and AHR immunostaining in sporadic GH-secreting adenomas. Immunostaining for AIP (upper panel) in four somatotroph adenomas (A, B, C and D), with the corresponding AHR immunostaining (lower panel). (A) A pure GH-secreting microadenoma treated with SSA before surgery, showing high AIP and AHR expression, with cytoplasmic (AHRc) and nuclear AHR (AHRn) localisation; (B) a pure, enclosed, GH-secreting, macroadenoma showing patchy AIP and AHRc expression; (C) a mixed, invasive, Gsp+ GH/PRL-secreting macroadenoma treated with SSA before surgery, showing diffuse AIP immunostaining with a subset of highly positive cells and exclusive AHRc immunostaining; (D) a huge, invasive, GH-secreting macroadenoma showing negative immunostaining for either AIP or AHR. The lowest panel (controls) shows AIP and AHR immunostaining in the normal pituitary (NP) tissue adjacent to a pituitary microadenoma (m, a microprolactinoma), where both AIP and AHR were barely detectable.
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
AIP and AHR immunostaining in sporadic GH-secreting adenomas. Immunostaining for AIP (upper panel) in four somatotroph adenomas (A, B, C and D), with the corresponding AHR immunostaining (lower panel). (A) A pure GH-secreting microadenoma treated with SSA before surgery, showing high AIP and AHR expression, with cytoplasmic (AHRc) and nuclear AHR (AHRn) localisation; (B) a pure, enclosed, GH-secreting, macroadenoma showing patchy AIP and AHRc expression; (C) a mixed, invasive, Gsp+ GH/PRL-secreting macroadenoma treated with SSA before surgery, showing diffuse AIP immunostaining with a subset of highly positive cells and exclusive AHRc immunostaining; (D) a huge, invasive, GH-secreting macroadenoma showing negative immunostaining for either AIP or AHR. The lowest panel (controls) shows AIP and AHR immunostaining in the normal pituitary (NP) tissue adjacent to a pituitary microadenoma (m, a microprolactinoma), where both AIP and AHR were barely detectable.
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
AIP and AHR immunostaining in sporadic GH-secreting adenomas. Immunostaining for AIP (upper panel) in four somatotroph adenomas (A, B, C and D), with the corresponding AHR immunostaining (lower panel). (A) A pure GH-secreting microadenoma treated with SSA before surgery, showing high AIP and AHR expression, with cytoplasmic (AHRc) and nuclear AHR (AHRn) localisation; (B) a pure, enclosed, GH-secreting, macroadenoma showing patchy AIP and AHRc expression; (C) a mixed, invasive, Gsp+ GH/PRL-secreting macroadenoma treated with SSA before surgery, showing diffuse AIP immunostaining with a subset of highly positive cells and exclusive AHRc immunostaining; (D) a huge, invasive, GH-secreting macroadenoma showing negative immunostaining for either AIP or AHR. The lowest panel (controls) shows AIP and AHR immunostaining in the normal pituitary (NP) tissue adjacent to a pituitary microadenoma (m, a microprolactinoma), where both AIP and AHR were barely detectable.
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
Factors influencing AIP–IHC in a series of 62 sporadic somatotropinomas and in a subgroup of 25 untreated cases
| All sporadic somatotropinomas (n=62) | Untreated only (n=25) | |||||
|---|---|---|---|---|---|---|
| High AIP–IHC | Low AIP–IHC | P | High AIP–IHC | Low AIP–IHC | P | |
| Patients (n) | 45 (72.6%) | 17 (27.4%) | 14 (56.0%) | 11 (44.0%) | ||
| Gender | 20 M/25 F | 5 M/12 F | 0.282 | 6 M/8 F | 3 M/8 F | 0.420 |
| Age (years) | 48 (18.0–78.0) | 42 (21.0–59.0) | 0.172 | 42.5 (18.0–64.0) | 41.0 (21.0–53.0) | 0.460 |
| Macroadenomas | 36/45 (77.8%) | 16/17 (94.1%) | 0.177 | 11/14 (78.6%) | 10/11 (90.9%) | 0.404 |
| Maximum diameter (mm) | 15.0 (8.0–40.0) | 20.5 (8.0–40.0) | 0.039 | 12.0 (8.0–40.0) | 22.0 (8.0–37.0) | 0.111 |
| Suprasellar extension | 15/43 (34.9%) | 14/16 (87.5%) | 0.0003 | 4/12 (33.3%) | 9/11 (81.8%) | 0.019 |
| Invasive tumours | 20/45 (44.4%) | 11/17 (64.7%) | 0.150 | 4/14 (28.6%) | 8/13 (61.5%) | 0.028 |
| Preoperative GH (ng/ml) | 5.6 (0.6–104.9) | 24.0 (1.2–82.0) | 0.006 | 11.9 (3.1–60.5) | 25.0 (4.6–32.0) | 0.156 |
| Preoperative IGF1 (%ULN) | 152.4 (38.5–611.6) | 226.0 (75.3–334.0) | 0.022 | 158.8 (51.3–611.6) | 221.2 (159.1–283.0) | 0.189 |
| Preoperative PRL (ng/ml) | 10.6 (3.8–174.0) | 15.1 (6.7–80.0) | 0.374 | 20.7 (7.6–92.5) | 22.1 (6.7–55.8) | 0.850 |
| Hyperprolactinemia (≥30 ng/ml) | 12/45 (26.7%) | 8/17 (47.1%) | 0.125 | 4/14 (28.6%) | 6/11 (54.5%) | 0.187 |
| IHC | ||||||
| GH | 27/45 | 8/17 | 0.297 | 9/14 | 5/11 | 0.607 |
| GH/PRL | 14/45 | 5/17 | 3/14 | 3/11 | ||
| Other mixed | 4/45 | 4/17 | 2/14 | 3/11 | ||
| Ki67 (%) | 1.0 (0.0–5.0) | 2.0 (0.0–8.0) | 0.019 | 1.1 (0.4–5.0) | 2.5 (0.0–8.0) | 0.130 |
| Gsp+ | 11/29 (37.9%) | 3/10 (30.0%) | 0.652 | 3/9 (33.3%) | 2/7 (28.6%) | 0.838 |
| Preoperative SSA treatment | 31/45 (68.9%) | 6/17 (35.3%) | 0.016 | – | – | – |
| Preoperative SSA treatment duration (months) | 7.5 (3.0–108.0) | 6.0 (4.5–38.0) | 0.851 | – | – | – |
IHC, immunohistochemistry; low and high AIP–IHC, AIP immunostaining score ≤2 and ≥3 respectively; Gsp+, tumours with somatic Gsp mutations.
Low AIP–IHC was associated with disease aggressiveness. All low AIP–IHC tumours but one were macroadenomas. Suprasellar extension was more frequent in low AIP–IHC vs high AIP–IHC tumours (P=0.0003), and this was confirmed in untreated cases (P=0.019). Low AIP–IHC tumours also tended to be more invasive than those retaining AIP expression, although the difference reached significance in untreated tumours only (P=0.028). A higher Ki67 index was found in invasive vs non-invasive GH-PA in the untreated group only (2.5% (1.0–8.0) vs 1.0% (0.0–3.0), P=0.043). Accordingly, a significant negative correlation was found between the AIP score and the Ki67 index in the whole series (P=0.011) and in untreated tumours (P=0.036), but not in SSA-treated tumours (Table 3). Significantly higher preoperative plasma GH, IGF1 (%ULN) and Ki67 index values were observed in low AIP–IHC vs high AIP–IHC tumours (P=0.006, P=0.022 and P=0.019 vs high AIP–IHC respectively), with similar non-significant trends in the untreated group (Table 2). Introducing preoperative SSA treatment as a covariant for each parameter, a significant negative association was confirmed between low AIP–IHC and higher Ki67 values (P=0.047), with a similar trend for higher preoperative plasma GH (P=0.065), but not for IGF1.
Correlations between AIP and the Ki67 and AHR scores in sporadic somatotropinomas
| All cases (n=53) | Untreated (n=23) | Treated (n=30) | ||||
|---|---|---|---|---|---|---|
| P | ρ | P | ρ | P | ρ | |
| AIP vs Ki67 | 0.011* | −0.35 | 0.036* | −0.47 | 0.362 | −0.16 |
| AIP vs AHRt | 0.048* | 0.27 | 0.050 | 0.41 | 0.341 | 0.18 |
| AIP vs AHRc | 0.012* | 0.35 | 0.077 | 0.38 | 0.101 | 0.30 |
| AIP vs AHRn | 0.478 | 0.10 | 0.497 | 0.14 | 0.667 | 0.08 |
| AHRc vs AHRn | 0.004* | 0.40 | 0.236 | 0.25 | 0.001* | 0.59 |
AHRt, total AHR score; AHRc, cytoplasmic AHR score; AHRn, nuclear AHR score. Results are given for ex-aequo correction. *Significant P values (<0.05).
Some AHRc immunostaining was also observed in most tumours with available data (50 out of 53 cases; 94.3%), including 29 cases (54.7%) that had a high AHRc score. AHRn was detected in nearly half of the studied cases (26 out of 53, 49.1%), of which 18 (34.0%) had a high AHRn score. Correlations between AIP and AHR immunoscores are shown in Table 3. Overall, a significant positive correlation was found between the AHRn and AHRc scores (P=0.004) whereas AIP correlated significantly with AHRc (P=0.012) and AHRt (P=0.049), but not with AHRn. Correlations between AIP and AHR approached significance in untreated tumours (P=0.05 for AHRt and P=0.077 for AHRc). By contrast, AHRn strongly correlated with AHRc in treated tumours only (P=0.0015).
Searching for factors able to influence AHR content, regardless of its intracellular localisation, we further analysed high AHR–IHC tumours, which accounted for 32.1% of the whole series (17/53). In high AHR–IHC tumours, microadenomas were more frequent (29.4 vs 8.3%, P=0.045), suprasellar extension was less frequent (25.0 vs 60.0%, P=0.02) and pure GH-secreting PA predominated over mixed secreting GH-PA (76.5 vs 44.4%, P=0.029) when compared with low AHR–IHC. By contrast, no significant influence of gender, tumour invasiveness or preoperative hormone profile was observed. The AHRt score and the characteristics of high AHR–IHC tumours were not statistically influenced by preoperative SSA treatment (Table 4).
Characterisation of high AHR–IHC sporadic somatotropinomas
| All sporadic somatotropinomas (n=53) | High AHRt–IHC (n=17) | |||||
|---|---|---|---|---|---|---|
| High AHRt–IHC | Low AHRt–IHC | P | Untreated | Treated | P | |
| Patients (n) | 17/53 (32.1%) | 36/53 (67.9%) | 6/23 (26.1%) | 11/30 (36.7%) | 0.413 | |
| Gender | 6 M/11 F | 15 M/21 F | 0.658 | 3 M/3 F | 3 M/8 F | 0.345 |
| Age (years) | 40 (21.0–52.0) | 45 (18.0–78.0) | 0.172 | 36.0 (21.0–51.0) | 48.0 (28.0–57.0) | 0.145 |
| Macroadenomas | 12/17 (70.6%) | 33/36 (91.7%) | 0.045 | 4/6 (66.7%) | 8/11 (72.7%) | 0.793 |
| Maximum diameter (mm) | 15.0 (8.0–40.0) | 15.0 (8.0–40.0) | 0.502 | 20.0 (9.0–40.0) | 15.0 (8.0–31.0) | 0.497 |
| Suprasellar extension | 4/16 (25.0%) | 21/35 (60.0%) | 0.020 | 2/6 (33.3%) | 2/10 (20.0%) | 0.551 |
| Invasive tumours | 8/17 (47.1%) | 19/36 (52.8%) | 0.697 | 3/6 (50.0%) | 5/11 (45.4%) | 0.858 |
| Preoperative GH (ng/ml) | 6.9 (1.0–44.6) | 8.3 (0.6–104.9) | 0.672 | 15.7 (4.3–29.0) | 6.0 (1.0–44.5) | 0.075 |
| Preoperative IGF1 (%ULN) | 159.1 (52.6–576.0) | 182.2.0 (38.5–338.9) | 0.561 | 172.0 (142.9–576.0) | 144.8 (52.6–352.0) | 0.346 |
| Preoperative PRL (ng/ml) | 14.3 (3.8–50.6) | 13.2 (4.2–174.0) | 0.282 | 20.0 (9.0–33.0) | 8.5 (3.8–56.0) | 0.161 |
| Hyperprolactinemia (≥30 ng/ml) | 4/17 (23.5%) | 14/36 (38.9%) | 0.270 | 2/6 (33.3%) | 2/11 (18.2%) | 0.488 |
| IHC | 0.092 | |||||
| GH | 13/17 | 16/36 | 5/6 | 8/11 | 0.738 | |
| GH/PRL | 3/17 | 15/36 | 1/6 | 2/11 | ||
| Other mixed | 1/17 | 5/36 | 0/6 | 1/11 | ||
| Ki67 (%) | 0.8 (0.0–6.0) | 0.8 (0.0–8.0) | 0.887 | 1.2 (0.5–6.0) | 0.5 (0.0–3.1) | 0.050 |
| Gsp+ | 2/9 (22.2%) | 11/22 (50.0%) | 0.155 | 0/2 (0.0%) | 2/7 (28.6%) | 0.91 |
| Preoperative SSA treatment | 11/17 (64.7%) | 19/36 (52.8%) | 0.413 | – | – | – |
| Preoperative SSA treatment duration | 6.2 (4.0–23.0) | 7.5 (3.0–108.0) | 0.381 | – | – | – |
IHC, immunohistochemistry; low and high AHRt–IHC, AHR total score <4 and ≥4 respectively; Gsp+, tumours with somatic Gsp mutations.
Relationship between AIP/AHR expression and the clinical response to SSA
In order to evaluate the relationship between AIP/AHR expression and the effect of preoperative SSA treatment, treated patients were divided into controlled, partially controlled and uncontrolled based on preoperative IGF1 (%ULN) values. Results are summarised in Table 5.
Relationship between AIP/AHR expression and the preoperative response to somatostatin analogues in 33 sporadic somatotropinomas
| Controlled | Partially controlled | Uncontrolled | P1a | P2a | |
|---|---|---|---|---|---|
| Patients (n) | 10 | 13 | 10 | ||
| Gender | 3 M/7 F | 7 M/6 F | 5 M/5 F | 0.493 | 0.730 |
| Age (years) | 52.0 (30.0–72.0) | 38.0 (21.0–61.0) | 44.0 (26.0–78.0) | 0.411 | 0.891 |
| Macroadenomas | 8/10 (80%) | 10/13 (76.9%) | 10/10 (100%) | 0.272 | 0.109 |
| Suprasellar extension | 4/10 (40%) | 6/13 (46.1%) | 6/10 (60%) | 0.478 | 0.238 |
| Invasive | 5/10 (50%) | 7/13 (53.8%) | 6/10 (60%) | 0.902 | 0.678 |
| GH at diagnosis (ng/ml) | 13.0 (4.0–57.9) | 44.0 (6.9–107.0) | 24.1 (4.8–75.0) | 0.234 | 0.827 |
| IGF1 at diagnosis (%ULN) | 269.5 (127.0–636.0) | 279.0 (158.0–418.6) | 239.4 (187.0–393.9) | 0.775 | 0.580 |
| Pretreatment duration (months) | 10.0 (4.0–108.0) | 5.0 (3.0–21.0) | 6.0 (4.5–23.0) | 0.303 | 0.887 |
| Preoperative GH (ng/ml) | 3.2 (0.6–7.0) | 5.3 (0.6–55.0) | 19.0 (1.1–104.9) | 0.053 | 0.028 |
| Preoperative IGF1 (%ULN) | 75.3 (38.5–104.0) | 165.6 (131.0–278.0) | 226.1 (144.8–352.0) | <0.0001 | <0.0001 |
| ΔGH (%) | −75.6 (−97.9; −46.1) | −65.6 (−97.1; −2.0) | −37.9 (−98.6; +66.7) | 0.082 | 0.031 |
| ΔIGF1 (%) | −66.5 (−88.6; −29.8) | −35.6 (−76.0; −29.6) | −27.9 (−13.1; +38.2) | 0.0004 | 0.0007 |
| Preoperative tumour shrinkage | 3/8 (37.5%) | 2/7 (28.6%) | 1/10 (10%) | 0.376 | 0.181 |
| Ki67 (%) | 0.5 (0.0–1.0) | 0.3 (0.0–3.0) | 0.8 (0.0–3.5) | 0.408 | 0.212 |
| High AIP–IHC | 9/10 (90%) | 13/13 (100%) | 5/10 (50%) | 0.006 | 0.002 |
| AIP score | 4.0 (1.0–6.0) | 4.0 (3.0–5.0) | 2.5 (1.0–5.0) | 0.014 | 0.008 |
| High AHR–IHC | 4/10 (40.0%) | 2/8 (25.0%) | 4/10 (40.0%) | 0.756 | 0.724 |
| AHRt score | 2.5 (1.0–6.0) | 2.0 (1.0–4.0) | 3.0 (1.0–6.0) | 0.747 | 0.588 |
| AHRc score | 2.0 (1.0–3.0) | 2.0 (1.0–3.0) | 2.0 (1.0–3.0) | 0.627 | 0.757 |
| AHRn score | 0.0 (0.0–3.0) | 0.0 (0.0–2.0) | 1.0 (0.0–3.0) | 0.642 | 0.203 |
P1, three groups' comparison by Kruskal–Wallis test; P2, two groups' comparison (controlled and partially controlled taken together vs uncontrolled) by Mann–Whitney U test.
For the comparison of immunoscores, results are given for ex-aequo correction.
By definition, preoperative IGF1 (%ULN) significantly differed among the three groups (P<0.0001). Differences in disease control were related to the effect of treatment, as plasma GH and IGF1 (%ULN) were similar at diagnosis. Accordingly, ΔIGF1 (P=0.0004) significantly differed among the three groups, with a similar but not significant trend for ΔGH (P=0.082). Uncontrolled cases had significantly lower ΔIGF1 (P=0.0007) and ΔGH (P=0.031) than controlled and partially controlled tumours taken together. Tumour shrinkage was reported in a subset of tumours (7/26, 26.7%).
The proportion of low AIP–IHC tumours and the AIP score significantly differed among the three groups (P=0.006 and P=0.014 respectively). In particular, uncontrolled tumours had a significantly lower AIP expression than controlled and partially controlled tumours taken together (low AIP–IHC in 50 vs 4.3% and median AIP score 2.5 vs 4.0, P=0.0003 and P=0.008 respectively). High AIP–IHC was observed in all tumours showing preoperative shrinkage, contrasting with low AIP–IHC in 6 out of 19 tumours in the absence of shrinkage (31.6%). Accordingly, the AIP score was significantly higher in the presence (4.5 (3–5)) than in the absence (3.0 (1–5)) of tumour shrinkage (P=0.014). In contrast, no significant difference in AHR expression was found that was attributable to the effect of preoperative SSA treatment in terms of IGF1 normalisation or tumour shrinkage.
The relationship between AIP–IHC and the response to SSA was also studied in tumours with proven normal AIP sequences. Half had received preoperative SSA (20/40 cases), of which seven showed a partial and seven a complete hormone response (i.e. 14 out of 20 responders), and tumour shrinkage was reported in 2 out of 15 cases. As shown in Fig. 2, the AIP score was significantly higher in the presence of a partial or complete hormone response (P=0.008 vs uncontrolled tumours) and in the presence of preoperative hormone shrinkage (P=0.046). AHR immunostaining was unrelated to the outcome of treatment, although a trend towards a higher AHRc score in controlled and partially controlled tumours was observed when compared with uncontrolled tumours (P=0.07) (data not shown).
Variations in the AIP immunoscore according to the preoperative response to somatostatin analogues in 20 sporadic somatotropinomas (normal AIP sequence subgroup). The AIP score was significantly higher in the presence of a partial or complete hormone response (**P=0.008, panel A) and in the presence of preoperative shrinkage (*P=0.046, panel B).
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
Variations in the AIP immunoscore according to the preoperative response to somatostatin analogues in 20 sporadic somatotropinomas (normal AIP sequence subgroup). The AIP score was significantly higher in the presence of a partial or complete hormone response (**P=0.008, panel A) and in the presence of preoperative shrinkage (*P=0.046, panel B).
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
Variations in the AIP immunoscore according to the preoperative response to somatostatin analogues in 20 sporadic somatotropinomas (normal AIP sequence subgroup). The AIP score was significantly higher in the presence of a partial or complete hormone response (**P=0.008, panel A) and in the presence of preoperative shrinkage (*P=0.046, panel B).
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
In vitro effect of octreotide on GH-PA
In order to search for a direct effect of octreotide on AIP and AHR expression in GH-PA, primary cultures from six GH-PA – one Gsp+ and five Gsp− – were treated with 10 nM octreotide in vitro. Data are illustrated in Fig. 3. Two GH-PA (both Gsp−) were suitable for time-dependent experiments, showing a significant increase in AIP expression after 24 h of treatment. The remaining tumours were studied at baseline and after 24 h of treatment. Measurable levels of AIP were detected in all tumours, whereas in two cases, no reliable quantification of AHR could be obtained. A variable but significant increase in AIP expression was observed in all cases (median 2.03-fold; range 1.2–13.9, P=0.027), whereas AHR expression was increased in a single Gsp− tumour only (median 1.2-fold, range 1.0–2.8, P, NS). The Gsp+ tumour showed a 2.9-fold increase in AIP expression and no increase in AHR.
In vitro effect of octreotide treatment on AIP and AHR expression for GH-secreting adenomas. Western blotting experiments were performed after protein extraction from primary cultures of six GH-secreting adenomas, incubated in the presence or in the absence of 10 nM octreotide. AIP:GADPH and AHR:GADPH ratios were obtained by densitometry of the relative bands. (A) A maximal increase in AIP expression was observed at 24 h in two GH-PA treated for up to 48 h (P=0.002 vs basal expression). (B) Shown are pooled data obtained for GH-PA after 24 h of octreotide treatment for AIP (right panel, n=6) and AHR (left panel, n=4) protein expression. Overall, a significant increase in AIP, but not in AHR, expression was observed (P=0.027). (C) Examples of western blotting results obtained at 24 h for Gsp− (right panel) and for Gsp+ (left panel) tumours. A significant increase in AIP, but not in AHR, was present in both cases.
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
In vitro effect of octreotide treatment on AIP and AHR expression for GH-secreting adenomas. Western blotting experiments were performed after protein extraction from primary cultures of six GH-secreting adenomas, incubated in the presence or in the absence of 10 nM octreotide. AIP:GADPH and AHR:GADPH ratios were obtained by densitometry of the relative bands. (A) A maximal increase in AIP expression was observed at 24 h in two GH-PA treated for up to 48 h (P=0.002 vs basal expression). (B) Shown are pooled data obtained for GH-PA after 24 h of octreotide treatment for AIP (right panel, n=6) and AHR (left panel, n=4) protein expression. Overall, a significant increase in AIP, but not in AHR, expression was observed (P=0.027). (C) Examples of western blotting results obtained at 24 h for Gsp− (right panel) and for Gsp+ (left panel) tumours. A significant increase in AIP, but not in AHR, was present in both cases.
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
In vitro effect of octreotide treatment on AIP and AHR expression for GH-secreting adenomas. Western blotting experiments were performed after protein extraction from primary cultures of six GH-secreting adenomas, incubated in the presence or in the absence of 10 nM octreotide. AIP:GADPH and AHR:GADPH ratios were obtained by densitometry of the relative bands. (A) A maximal increase in AIP expression was observed at 24 h in two GH-PA treated for up to 48 h (P=0.002 vs basal expression). (B) Shown are pooled data obtained for GH-PA after 24 h of octreotide treatment for AIP (right panel, n=6) and AHR (left panel, n=4) protein expression. Overall, a significant increase in AIP, but not in AHR, expression was observed (P=0.027). (C) Examples of western blotting results obtained at 24 h for Gsp− (right panel) and for Gsp+ (left panel) tumours. A significant increase in AIP, but not in AHR, was present in both cases.
Citation: Endocrine-Related Cancer 20, 5; 10.1530/ERC-12-0322
Influence of Gsp mutations on AIP and AHR expression
The characteristics of Gsp+ and Gsp− somatotropinomas are summarised in Table 6. No significant difference was observed at diagnosis between the two groups. Among tumours treated with SSA preoperatively (9 Gsp+ and 14 Gsp−), Gsp+ tumours tended to be associated with a higher reduction in plasma GH values and a higher rate of shrinkage before surgery (P=0.09 vs Gsp− for both parameters). Significantly, lower Ki67 values were also observed in Gsp+ when compared with Gsp− tumours in SSA-treated cases (P=0.006).
Clinical characteristics and AIP/AHR expression in 39 sporadic somatotroph adenomas with or without Gsp mutations
| Gsp+ | Gsp− | Pa | |
|---|---|---|---|
| Tumours (n) | 14 | 25 | |
| Gender | 6 M/8 F | 9 M/16 F | 0.673 |
| Age (years) | 51.5 (30.0–72.0) | 45.0 (21.0–61.0) | 0.147 |
| Macroadenomas | 13/14 (92.9%) | 19/25 (76.0%) | 0.188 |
| Suprasellar extension | 5/13 (38.5%) | 13/23 (56.5%) | 0.298 |
| Maximum diameter (mm) | 14.0 (8.0–30.0) (10) | 15.0 (8.0–40.0) (20) | 0.680 |
| Invasive | 5/14 (35.7%) | 13/25 (52.0%) | 0.328 |
| GH at diagnosis | 22.5 (4.0–56.9) (14) | 13.5 (4.0–107.0) (20) | 0.916 |
| IGF1 at diagnosis (%ULN) | 239.9 (51.3–678.0) (13) | 283.5 (96.1–814.0) (20) | 0.127 |
| PRL at diagnosis | 20.1 (6.7–152.8) (12) | 22.0 (3.2–66.0) (20) | 0.951 |
| Preoperative SSA | 9/14 (64.3%) | 14/25 (56.0%) | 0.614 |
| ΔGH (%)b | −77.6 (−95.2; −30.6) (8) | −56.0 (−94.4; −2.0) (12) | 0.093 |
| ΔIGF1 (%)b | −29.9 (−84.5; −25.7) (6) | −43.4 (−79.8; −20.0) (10) | 0.605 |
| Ki67 (%) | |||
| All | 0.5 (0.0–3.0) (11) | 1.2 (0.0–5.0) (22) | 0.061 |
| Treated | 0.0 (0.0–0.5) (7) | 1.2 (0.0–3.1) (13) | 0.006 |
| Untreated | 2.3 (1.0–3.0) (4) | 1.3 (0.0–5.0) (9) | 0.589 |
| Low AIP–IHC | |||
| All | 3/14 | 7/25 | 0.652 |
| Treated | 1/9 | 2/14 | 0.825 |
| Untreated | 2/5 | 5/11 | 0.838 |
| AIP score | |||
| All | 4.0 (2.0–5.0) | 4.0 (0.0–6.0) | 0.765 |
| Treated | 4.0 (2.0–5.0) | 4.0 (2.0–6.0) | 0.897 |
| Untreated | 3.0 (2.0–5.0) | 3.0 (0.0–5.0) | 0.645 |
| AHRt score | |||
| All | 2.0 (0.0–5.0) | 2.5 (0.0–6.0) | 0.259 |
| Treated | 2.0 (1.0–5.0) | 4.0 (2.0–6.0)* | 0.095 |
| Untreated | 2.0 (0.0–2.0) | 2.0 (0.0–4.0) | 0.517 |
| AHRc score | |||
| All | 2.0 (0.0–3.0) | 2.0 (0.0–3.0) | 0.415 |
| Treated | 2.0 (1.0–3.0) | 2.0 (1.0–3.0)* | 0.999 |
| Untreated | 1.5 (0.0–2.0) | 1.0 (0.0–2.0) | 0.705 |
| AHRn score | |||
| All | 0.0 (0.0–2.0) | 2.0 (0.0–3.0) | 0.025 |
| Treated | 0.0 (0.0–2.0) | 2.0 (0.0–3.0) | 0.031 |
| Untreated | 0.0 (0.0–1.0) | 1.0 (0.0–2.0) | 0.227 |
SSA, somatostatin analogues; Gsp+, Gsp mutation; Gsp−, no Gsp mutation. *P<0.05 vs untreated Gsp− adenomas.
For the comparison of immunoscores, all results are given for ex-aequo correction.
Treated cases only.
No significant effect of Gsp status was observed on AIP expression. In particular, as observed in unselected GH-PA, a lower AIP score was observed in Gsp+ tumours in the presence of a suprasellar extension (P=0.034 vs intrasellar Gsp+ tumours, data nor shown). The only effect of Gsp status was a significantly lower AHRn score in Gsp+ tumours (P=0.025 vs Gsp− tumours), suggesting cytoplasmic retention of AHR in Gsp+ GH-PA. This finding was confirmed in the presence of preoperative SSA treatment (P=0.031 vs treated Gsp− tumours). In contrast, preoperative SSA treatment was associated with a significant increase in AHR expression in Gsp− tumours (P=0.013 for AHRc and P=0.037 for AHRt vs untreated Gsp− tumours respectively), with no significant change in its nuclear localisation.
Discussion
This extensive series of tumours from patients with sporadic acromegaly provides substantial new information on the effects of SSA on the expression of AIP and its best characterised molecular partner, AHR. In addition, it evaluates for the first time the influence of somatic Gsp mutations on such parameters.
Data obtained across the whole series are consistent with previous reports showing AIP down-regulation in aggressive somatotropinomas (Jaffrain-Rea et al. 2009, Kasuki Jomori de Pinho et al. 2011). However, tumour volume and proliferative activity were found to have a stronger impact on AIP expression than invasive features. AIP loss was significantly associated with suprasellar extension and higher Ki67 values regardless of preoperative SSA treatment, but it was significantly associated with invasiveness in untreated cases only. This indicates that preoperative pharmacological treatment introduces an important limitation in the use of AIP immunostaining as a marker of invasiveness in somatotropinomas. This is similar to the effect of SSA pretreatment on the Ki67 index, which can be significantly reduced by the anti-proliferative effects of SSA in GH-PA (Losa et al. 2001, Jaffrain-Rea et al. 2002, this study). Overall, low AIP expression was also associated with higher preoperative GH/IGF1 levels. Although this could, in part, be due to the lower proportion of low AIP–IHC tumours treated by SSA preoperatively, a trend towards higher preoperative GH levels in low AIP–IHC tumours was seen after correction for preoperative treatment on the whole series and in untreated tumours. Taken together, these findings indicate that low AIP–IHC tumours share common characteristics with AIPmut GH-PA (Daly et al. 2010). Notably, the higher percentage of GH/PRL adenomas reported in AIPmut PA has not been observed among low AIP–IHC somatotropinomas (Kasuki et al. 2012, this study). The mechanisms of AIP down-regulation in non-AIPmut somatotropinomas have been poorly investigated, but overexpression of the miR-107 has been recently proposed (Trivellin et al. 2012).
Dysregulation of AHR expression has been reported in a number of tumours (Harper et al. 2006, Dietrich & Kaina 2010) and reduced AHR (Jaffrain-Rea et al. 2009) and ARNT (Heliövaara et al. 2009, Raitila et al. 2010) expression have been observed in AIPmut PA. In this study, only 32% of sporadic GH-PA displayed a high AHR content. We found these tumours to be smaller and to include a higher proportion of pure GH-secreting PA than those displaying a low AHR content, suggesting that AHR down-regulation may also occur during the evolution of sporadic GH-PA. A significant correlation was found between AIP and AHRc, further supporting a role for AIP in the stabilisation of AHR in PA (Jaffrain-Rea et al. 2009). This may be attenuated by preoperative SSA treatment, potentially suggesting that the stability of the AIP/AHR complex may be influenced by SSA or that SSA differentially affect AIP and AHR expression in GH-PA. By contrast, despite AHR signalling being potentially enhanced by AIP (Petrulis & Perdrew 2002), AHRn was unrelated to AIP, irrespective of SSA treatment. Thus, the potential influence of AIP on AHR signalling in sporadic GH-PA remains unclear.
Data from a large series of sporadic GH-PA with proven normal AIP sequences in the majority of cases support recent evidence that AIP expression is increased by preoperative SSA treatment (Jaffrain-Rea et al. 2010, Chahal et al. 2012). Notably, similar results were obtained whether lanreotide (Jaffrain-Rea et al. 2010, Chahal et al. 2012, this study) or octreotide (Jaffrain-Rea et al. 2010, this study) was used for the preoperative treatment of acromegalic patients. In contrast, no change in AIP expression was observed in two AIPmut cases operated before and after treatment (Jaffrain-Rea et al. 2010). Supporting recent findings in GH3 cells (Chahal et al. 2012), this study provides the first evidence for a variable but significant up-regulation of AIP expression by octreotide in human GH-PA in vitro, with a median twofold increase. A significant relationship was also found between AIP immunostaining and the outcome of pre-surgical SSA treatment, both in terms of hormonal control and tumour shrinkage, further indicating AIP to be an important mediator of SSA in GH-PA. Indeed, the AIP score was significantly lower in uncontrolled tumours when compared with partially and fully controlled tumours and significantly higher in the presence of preoperative shrinkage. In particular, none of the tumours showing preoperative shrinkage had low AIP immunostaining, contrasting with 30% of those showing no effect of SSA on tumour volume. These results echo those obtained in a large series of AIPmut acromegalic patients, in which the median shrinkage with SSA was highly statistically significantly lower than in control acromegalic patients (0 vs 41%, P<0.000001; Daly et al. 2010). In this study, no difference was found according to the degree of disease control obtained at the time of surgery (IGF1 normalisation or reduction by ≥30% compared with baseline values). Owing to the retrospective and multicentre characteristics of our study, we might have been unable to distinguish between these two groups. Preoperative SSA treatment was not standardised, so that differences in the degree of response obtained before surgery did not necessarily reflect differences in SSA sensitivity. Additional factors may also be involved in the modulation of AIP by SSA. Chahal et al. (2012) found a positive correlation between AIP expression and IGF1 changes in female acromegalics only. A similar non-significant trend was present in our series (data not shown), but the higher proportion of female cases could have introduced some bias and no gender specificity could be shown. However, the unusual male predominance in AIPmut somatotropinomas (Daly et al. 2010) and reported interactions between AIP, AHR and steroid receptors (Beischlag et al. 2008), in particular oestrogen receptor α (Matthews & Gustafsson 2006, Cai et al. 2011), suggest potential variations in AIP expression or function according to patient's gender and the steroid milieu, which deserve further investigation.
Because the cyclin kinase inhibitor p27Kip1 can be induced by SSA (Ferrante et al. 2006) and by AHR signalling (Marlowe & Puga 2005), we proposed the hypothesis that AHR itself could be a target of SSA in GH-PA. However, no significant difference in AHR content or in its nuclear localisation was observed between SSA-treated vs -untreated tumours, or according to the effect of treatment. Although there was some evidence that AHRc immunostaining could be higher in SSA-treated tumours, this was also unrelated to the effect of treatment and might be due to the higher AIP expression observed in treated GH-PA. Unexpectedly, a strong correlation was found between AHRn and AHRc in treated tumours. This may reflect the complex regulation of nucleocytoplasmic shuttling of AHR, which involves several exogenous and endogenous activators (Beischlag et al. 2008, Nguyen & Bradfield 2008). Overall, these data argue against a significant role for AHR in the pharmacological response to SSA in GH-PA.
Most of the pharmacological effects of octreotide and lanreotide on GH-PA are mediated by somatostatin receptor (SSTR) subtypes 2 and 5 (Ben-Shlomo & Melmed 2010). AIP immunostaining was identified as a predictive factor of post-operative SSA sensitivity independent of SSTR2 (Kasuki et al. 2012) and a slightly higher SSTR5 expression was observed in AIPmut tumours (Chahal et al. 2012), suggesting that AIP-related differences in SSA sensitivity are not due to important changes in SSTRs. Potential variations in the expression of receptors reported to negatively affect the pharmacological response to SSA, such as truncated variants of SSTR5 (Durán-Prado et al. 2010, 2012) or the dopamine receptor D2R (Zatelli et al. 2005), have not been reported yet. An alternative molecular link between AIP expression and SSA signalling is ZAC1, a tumour suppressor gene involved in the anti-proliferative action of octreotide (Theodoropoulou et al. 2006), whose expression correlates with tumour shrinkage and IGF1 changes in treated GH-PA (Theodoropoulou et al. 2009). ZAC1 has been recently proposed to act downstream of AIP in GH3 cells (Chahal et al. 2012). However, potential correlations between AIP and ZAC1 expression in human GH-PA remain to be further explored.
A relationship between AIP, AHR and cAMP signalling, the best characterised pathway in GH-PA and an important target of SSA, is suggested by experimental evidence indicating that i) cAMP is a non-ligand activator of AHR able to induce transcriptional responses different from exogenous ligands (Oesch-Bartlomowicz et al. 2005), ii) nucleocytoplasmic shuttling of AHR induced by cAMP is inhibited by PDE2, which is stabilised by AIP (de Oliveira et al. 2007) and iii) AIP interactions with PDE4A5 can be disrupted by AIP mutations (Leontiou et al. 2008). There is also very recent evidence that AIP reduces forskolin-induced cAMP signalling in GH3 cells, although forskolin did not influence AIP expression (Formosa et al. 2013). However, intracellular cAMP concentrations have not been reported in the presence of AIP abnormalities in GH-PA and the potential effect of cAMP signalling on AIP expression in these tumours is unknown. Because Gsp mutations induce a constitutive activation of the cAMP/PKA pathway, we studied AIP and AHR expression according to Gsp status. In agreement with most reports, no specific phenotype was observed in Gsp+ tumours, although they tended to respond better than Gsp− tumours to preoperative SSA treatment (Barlier et al. 1998, Lania et al. 2003). Accordingly, treated Gsp+ tumours had significantly lower Ki67 than their Gsp− counterpart. AIP expression was similar in Gsp+ and Gsp− tumours. Because only large Gsp+ tumours had a low AIP expression, down-regulation of AIP appears as a late event in Gsp+-related tumorigenesis. Notably, no Gsp mutations were found in a small series of AIPmut GH-PA, further suggesting that these are independent pathogenetic events (Angelini et al. 2010). Preoperative SSA treatment had no differential effect on AIP expression in Gsp+ and Gsp− tumours, and the effect of octreotide on AIP expression was unremarkable in the Gsp+ tumour studied in vitro. In contrast, some effect of Gsp status could be found on AHR expression and localisation. In Gsp− tumours, preoperative SSA was associated with a significantly higher AHR content and AHR was increased by octreotide in one of three Gsp− tumours studied in vitro. Therefore, SSA may induce AHR in a subset of Gsp− GH-PA. Unexpectedly, AHRn expression was lower in Gsp+ vs Gsp− GH-PA, regardless of AHR content, suggesting cytoplasmic retention of AHR in Gsp+ tumours. A potential explanation is that Gsp+ tumours activate regulatory mechanisms that operate to counteract the increase in cAMP concentration (Lania et al. 2003, Pertuit et al. 2009), among which is an increased PDE activity, which might in turn inhibit AHRn shuttling. In addition, cAMP signalling involves different subcellular compartments, which may differentially regulate endocrine functions and AHR activation in somatototroph cells. Indeed, cytoplasmic retention of AHR was also observed in treated Gsp+ tumours. The biological implications of these findings remain to be further investigated.
In conclusion, this study further supports a role for AIP down-regulation in the pathogenesis of sporadic acromegaly, with low AIP-expressing tumours sharing phenotypic features with AIPmut somatotropinomas. Overall, AIP down-regulation is associated with a reduced AHRc expression, with no significant effect on its nuclear localisation. AHR, but not AIP, appears to be differentially regulated according to Gsp status. In contrast, AIP, but not AHR, is an important mediator of SSA in GH-PA.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was partially supported by grants from the Italian Ministry for University and Research (MIUR), the ‘Carlo Ferri’ Foundation for prevention in oncology, Monterotondo (RM), Italy and by the Fonds d'Investissement pour la Recherche Scientifique (FIRS) du CHU de Liège, Belgium.
References
Angelini M, Turchi A, Tichomirowa M, Esposito V, Alesse E, Beckers A & Jaffrain-Rea ML. Somatotropinomas in patients with germline AIP mutations: absence of somatic mutations in the GNAS1 gene. 15th Congress of the European Neuroendocrine Association, Liège, 22–25 September 2010, Belgium (p 148)
Barlier A, Gunz G, Zamora AJ, Morange-Ramos I, Figarella-Branger D, Dufour H, Enjalbert A & Jaquet P 1998 Pronostic and therapeutic consequences of Gsα mutations in somatotroph adenomas. Journal of Clinical Endocrinology and Metabolism 83 1604–1610. (doi:10.1210/jc.83.5.1604).
Barlier A, Vanbellinghen JF, Daly AF, Silvy M, Jaffrain-Rea ML, Trouillas J, Tamagno G, Laure Cazabat L, Bours V & Brue T et al. 2007 Mutations in the aryl hydrocarbon receptor interacting protein gene are not highly prevalent among subjects with sporadic pituitary adenomas. Journal of Clinical Endocrinology and Metabolism 92 1952–1955. (doi:10.1210/jc.2006-2702).
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).
Beischlag TV, Luis Morales J, Hollingshead BD & Perdew GH 2008 The aryl hydrocarbon receptor complex and the control of gene expression. Critical Reviews in Eukaryotic Gene Expression 18 207–250. (doi:10.1615/CritRevEukarGeneExpr.v18.i3.20).
Ben-Shlomo A & Melmed S 2010 Pituitary somatostatin receptor signaling. Trends in Endocrinology and Metabolism 21 123–133. (doi:10.1016/j.tem.2009.12.003).
Bolger GB, Peden AH, Steele MR, MacKenzie C, McEwan DG, Wallace DA, Huston E, Baillie GS & Houslay MD 2003 Attenuation of the activity of the cAMP-specific phosphodiesterase PDE4A5 by interaction with the immunophilin XAP2. Journal of Biological Chemistry 278 33351–33363. (doi:10.1074/jbc.M303269200).
Cai W, Kramarova TV, Berg P, Korbonits M & Pongratz I 2011 The immunophilin-like protein XAP2 is a negative regulator of estrogen signaling through interaction with estrogen receptor α. PLoS ONE 6 e25201. (doi:10.1371/journal.pone.0025201).
Chahal HS, Trivellin G, Leontiou CA, Alband N, Fowkes RC, Tahir A, Igreja SC, Chapple JP, Jordan S & Lupp A et al. 2012 Somatostatin analogs modulate AIP in somatotroph adenomas: the role of the ZAC1 pathway. Journal of Clinical Endocrinology and Metabolism 97 E1411–E1420. (doi:10.1210/jc.2012-1111).
Daly AF, Vanbellinghen JF, Khoo SK, Jaffrain-Rea ML, Naves LA, Guitelman MA, Murat A, Emy P, Gimenez-Roqueplo AP & Tamburrano G et al. 2007 Aryl hydrocarbon receptor-interacting protein gene mutations in familial isolated pituitary adenomas: analysis in 73 families. Journal of Clinical Endocrinology and Metabolism 92 1891–1896. (doi:10.1210/jc.2006-2513).
Daly AF, Tichomirowa MA, Petrossians P, Heliövaara E, Jaffrain-Rea ML, Barlier A, Naves LA, Ebeling T, Karhu A & Raappana A et al. 2010 Clinical characteristics and therapeutic responses in patients with germ-line AIP mutations and pituitary adenomas: an international collaborative study. Journal of Clinical Endocrinology and Metabolism 95 E373–E383. (doi:10.1210/jc.2009-2556).
Dietrich C & Kaina B 2010 The aryl hydrocarbon receptor (AhR) in the regulation of cell–cell contact and tumor growth. Carcinogenesis 31 1319–1328. (doi:10.1093/carcin/bgq028).
Durán-Prado M, Saveanu A, Luque RM, Gahete MD, Gracia-Navarro F, Jaquet P, Dufour H, Malagón MM, Culler MD & Barlier A et al. 2010 A potential inhibitory role for the new truncated variant of somatostatin receptor 5, sst5TMD4, in pituitary adenomas poorly responsive to somatostatin analogs. Journal of Clinical Endocrinology and Metabolism 95 2497–2502. (doi:10.1210/jc.2009-2247).
Durán-Prado M, Gahete MD, Delgado-Niebla E, Martínez-Fuentes AJ, Vázquez-Martínez R, García-Navarro S, Gracia-Navarro F, Malagon MM, Luque RM & Castaño JP 2012 Truncated variants of pig somatostatin receptor subtype 5 (sst5) act as dominant-negative modulators for sst2-mediated signaling. American Journal of Physiology. Endocrinology and Metabolism 303 E1325–E1334. (doi:10.1152/ajpendo.00445.2012).
Ferrante E, Pellegrini C, Bondioni S, Peverelli E, Locatelli M, Gelmini P, Lucani P, Peri A, Mantovani G & Bosari S et al. 2006 Octreotide promotes apoptosis in human somatotroph tumor cells by activating somatostatin receptor type 2. Endocrine-Related Cancer 13 955–962. (doi:10.1677/erc.1.01191).
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).
Harper PA, Riddick DS & Okey AB 2006 Regulating the key regulator: factors that control levels and activity of the aryl hydrocarbon receptor. Biochemical Pharmacology 72 267–279. (doi:10.1016/j.bcp.2006.01.007).
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).
Hubina E, Nanzer AM, Hanson MR, Ciccarelli E, Losa M, Gaia D, Papotti M, Terreni MR, Khalaf S & Jordan S et al. 2006 Somatostatin analogues stimulate p27 expression and inhibit the MAP kinase pathway in pituitary tumours. European Journal of Endocrinology 155 371–379. (doi:10.1530/eje.1.02213).
Iwata T, Yamada S, Mizusawa N, Golam HM, Sano T & Yoshimoto K 2007 The aryl hydrocarbon receptor-interacting protein gene is rarely mutated in sporadic GH-secreting adenomas. Clinical Endocrinology 66 499–502. (doi:10.1111/j.1365-2265.2007.02758.x).
Jaffrain-Rea ML, Di Stefano D, Minniti G, Esposito V, Bultrini A, Ferretti E, Santoro A, Faticanti Scucchi L, Gulino A & Cantore G 2002 A critical reappraisal of MIB-1 labelling index significance in a large series of pituitary tumours: secreting versus non-secreting adenomas. Endocrine-Related Cancer 9 103–113. (doi:10.1677/erc.0.0090103).
Jaffrain-Rea ML, Angelini M, Gargano D, Tichomirowa MA, Daly AF, Vanbellinghen JF, 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).
Jaffrain-Rea ML, Angelini M, Tichomirowa MA, Theodoropoulou M, Daly AF, Barlier A, Naves LA, Fajardo C, Zacharieva S, Rohmer V et al. 2010. Factors associated with AIP expression in somatotropinomas and the possible influence of somatostatin analogues. 15th Congress of the European Neuroendocrine Association, Liège, 22–25 September 2010, Belgium (p 78)
Kasuki Jomori de Pinho L, Vieira Neto L, Armondi Wildemberg LE, Gasparetto EL, Marcondes J, de Almeida Nunes B, Takiya CM & Gadelha MR 2011 Low aryl hydrocarbon receptor-interacting protein expression is a better marker of invasiveness in somatotropinomas than Ki-67 and p53. Neuroendocrinology 94 39–48. (doi:10.1159/000322787).
Kasuki L, Vieira Neto L, Wildemberg LE, Colli LM, de Castro M, Takiya CM & Gadelha MR 2012 AIP expression in sporadic somatotropinomas is a predictor of the response to octreotide LAR therapy independent of SSTR2 expression. Endocrine-Related Cancer 19 L25–L29. (doi:10.1530/ERC-12-0020).
Lania A, Persani L, Ballarè E, Mantovani S, Losa M & Spada A 1998 Constitutively active Gsα is associated with an increased phosphodiesterase activity in human growth hormone-secreting adenomas. Journal of Clinical Endocrinology and Metabolism 83 1624–1628. (doi:10.1210/jc.83.5.1624).
Lania AG, Mantovani G & Spada A 2003 Genetics of pituitary tumours: focus on G-protein mutations. Experimental Biology and Medicine 228 1004–1017.
Lania AG, Mantovani G, Ferrero S, Pellegrini C, Bondioni S, Peverelli E, Braidotti P, Locatelli M, Zavanone ML & Ferrante E et al. 2004 Proliferation of transformed somatotroph cells related to low or absent expression of protein kinase A regulatory subunit 1A protein. Cancer Research 64 9193–9198. (doi:10.1158/0008-5472.CAN-04-1847).
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).
Losa M, Ciccarelli E, Mortini P, Barzaghi R, Gaia D, Faccani G, Papotti M, Mangili F, Terreni MR & Camanni F et al. 2001 Effects of octreotide treatment on the proliferation and apoptotic index of GH-secreting pituitary adenomas. Journal of Clinical Endocrinology and Metabolism 86 5194–5200. (doi:10.1210/jc.86.11.5194).
Marlowe JL & Puga A 2005 Aryl hydrocarbon receptor, cell cycle regulation, toxicity, and tumorigenesis. Journal of Cellular Biochemistry 96 1174–1184. (doi:10.1002/jcb.20656).
Matthews J & Gustafsson JA 2006 Estrogen receptor and aryl hydrocarbon receptor signaling pathways. Nuclear Receptor Signaling 4 e016. (doi:10.162/nrs.04016).
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).
Oesch-Bartlomowicz B, Huelster A, Wiss O, Antoniou-Lipfert P, Dietrich C, Arand M, Weiss C, Bockamp E & Oesch F 2005 Aryl hydrocarbon receptor activation by cAMP vs. dioxin: divergent signaling pathways. PNAS 102 9218–9223. (doi:10.1073/pnas.0503488102).
Occhi G, Jaffrain-Rea ML, Trivellin G, Albiger N, Ceccato F, De Menis E, Angelini M, Ferasin S, Beckers A & Mantero F et al. 2010 The R304X mutation of the aryl hydrocarbon receptor interacting protein gene in familial isolated pituitary adenomas: mutational hot-spot or founder effect? Journal of Endocrinological Investigation 33 800–805. (doi:10.3275/6956).
Occhi G, Losa M, Albiger N, Trivellin G, Regazzo D, Scanarini M, Monteserin-Garcia JL, Fröhlich B, Ferasin S & Terreni MR et al. 2011 The glucose-dependent insulinotropic polypeptide receptor is overexpressed amongst GNAS1 mutation-negative somatotropinomas and drives growth hormone (GH)-promoter activity in GH3 cells. Journal of Neuroendocrinology 23 641–649. (doi:10.1111/j.1365-2826.2011.02155.x).
de Oliveira SK, Hoffmeister M, Gambaryan S, Müller-Esterl W, Guimaraes JA & Smolenski AP 2007 Phosphodiesterase 2A forms a complex with the co-chaperone XAP2 and regulates nuclear translocation of the aryl hydrocarbon receptor. Journal of Biological Chemistry 282 13656–13663. (doi:10.1074/jbc.M610942200).
Pertuit M, Barlier A, Enjalbert A & Gérard C 2009 Signalling pathway alterations in pituitary adenomas: involvement of Gsα, cAMP and mitogen-activated protein kinases. Neuroendocrinology 21 869–877. (doi:10.1111/j.1365-2826.2009.01910.x).
Petrulis JR & Perdrew GH 2002 The role of chaperone proteins in the aryl hydrocarbon receptor core complex. Chemico-Biological Interactions 141 25–40. (doi:10.1016/S0009-2797(02)00064-9).
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).
Theodoropoulou M, Zhang J, Laupheimer S, Paez-Pereda M, Erneux C, Florio T, Pagotto U & Stalla GK 2006 Octreotide, a somatostatin analogue, mediates its antiproliferative action in pituitary tumor cells by altering phosphatidylinositol 3-kinase signaling and inducing Zac1 expression. Cancer Research 66 1576–1582. (doi:10.1158/0008-5472.CAN-05-1189).
Theodoropoulou M, Tichomirowa MA, Sievers C, Yassouridis A, Arzberger T, Hougrand O, Deprez M, Daly AF, Petrossians P & Pagotto U et al. 2009 Tumor ZAC1 expression is associated with the response to somatostatin analog therapy in patients with acromegaly. International Journal of Cancer 125 2122–2126. (doi:10.1002/ijc.24602).
Trivellin G, Butz H, Delhove J, Igreja S, Chahal HS, Zivkovic V, McKay T, Patócs A, Grossman AB & Korbonits M 2012 MicroRNA miR-107 is overexpressed in pituitary adenomas and in vitro inhibits the expression of aryl hydrocarbon receptor-interacting protein (AIP). American Journal of Physiology. Endocrinology and Metabolism 303 E708–E719. (doi:10.1152/ajpendo.00546.2011).
Vierimaa O, Georgitsi M, Lehtonen R, Vahteristo P, Kokko A, Raitila A, Tuppurainen K, Ebeling TM, 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).
Zatelli MC, Piccin D, Tagliati F, Bottoni A, Ambrosio MR, Margutti A, Scanarini M, Bondanelli M, Culler MD & degli Uberti EC 2005 Dopamine receptor subtype 2 and somatostatin receptor subtype 5 expression influences somatostatin analogs effects on human somatotroph pituitary adenomas in vitro. Journal of Molecular Endocrinology 35 333–341. (doi:10.1677/jme.1.01876).
