Abstract
Primary pigmented nodular adrenocortical disease (PPNAD), whether in the context of Carney complex (CNC) or isolated, leads to ACTH-independent Cushing's syndrome (CS). CNC and PPNAD are caused typically by inactivating mutations of PRKAR1A, a gene coding for the type 1a regulatory subunit (R1α) of cAMP-dependent protein kinase (PKA). Mice lacking Prkar1a, specifically in the adrenal cortex (AdKO) developed CS caused by bilateral adrenal hyperplasia (BAH), which is formed from the abnormal proliferation of fetal-like adrenocortical cells. Celecoxib is a cyclooxygenase 2 (COX2) inhibitor. In bone, Prkar1a inhibition is associated with COX2 activation and prostaglandin E2 (PGE2) production that, in turn, activates proliferation of bone stromal cells. We hypothesized that COX2 inhibition may have an effect in PPNAD. In vitro treatment of human cell lines, including one from a patient with PPNAD, with celecoxib resulted in decreased cell viability. We then treated AdKO and control mice with 1500 mg/kg celecoxib or vehicle. Celecoxib treatment led to decreased PGE2 and corticosterone levels, reduced proliferation and increased apoptosis of adrenocortical cells, and decreased steroidogenic gene expression. We conclude that, in vitro and in vivo, celecoxib led to decreased steroidogenesis. In a mouse model of PPNAD, celecoxib caused histological changes that, at least in part, reversed BAH and this was associated with a reduction of corticosterone levels.
Introduction
Primary pigmented nodular adrenocortical disease (PPNAD) leads to adrenocorticotropin hormone (ACTH)-independent Cushing's syndrome (CS) (Carney et al. 1985, Stratakis 2007) and is caused typically by inactivating mutations of PRKAR1A, a gene coding for the type 1a regulatory subunit (R1α) of cAMP-dependent protein kinase (PKA) (Kirschner et al. 2000, Stratakis 2008). Prkar1a mice with the gene entirely knocked out (KO) specifically in the adrenal cortex (AdKO) developed bilateral adrenal hyperplasia (BAH) that formed from resurgence and proliferation of fetal-like adrenocortical cells (FLACs) (Sahut-Barnola et al. 2010, de Joussineau et al. 2012). These mice developed a late-onset CS that was more prominent in female animals, as is true for humans with PPNAD (Bertherat et al. 2009). FLACs have since been found in the adrenal cortex of humans with ACTH-independent CS and various forms of BAH (Carney et al. 2010, 2011, 2012) and are currently considered responsible for PPNAD and related disorders (Almeida & Stratakis 2011).
Similarly, in bone, cells of multipotent nature were found to be responsible for yet another benign neoplasm that is occasionally associated with PPNAD in the context of Carney complex (CNC), osteochondromyxoma (Carney et al. 2001). In both the Prkar1a heterozygous mouse (Kirschner et al. 2005) and in the double heterozygous mice deficient in both Prkar1a and Prkaca (Prkar1a+/−Prkaca+/−) (Tsang et al. 2010), the main catalytic subunit of PKA, abnormal proliferation of adult bone stromal cells (aBSCs) led to multiple tumors and fibrous dysplasia-like (FDL) lesions. In both these animal models, aBSC neoplastic growth was fueled by prostaglandin E2 (PGE2; Almeida et al. 2011). PGE2 is known to maintain stem cell-like cellular populations in various tissues (Cha et al. 2006, Greenhough et al. 2009, Daikoku et al. 2014).
Celecoxib, an FDA-approved drug for use in humans and effective inhibitor of the prostaglandin-endoperoxide synthase 2 (PTGS2) enzyme (also known as cyclooxygenase 2 (COX2)) has been used with the intent to decrease neoplastic growth and/or proliferation in a variety of tumors (Bertagnolli 2006, Daikoku et al. 2014, Yokouchi et al. 2014, Kurtova et al. 2015). We hypothesized that celecoxib might also be effective in decreasing the proliferation of adrenocortical cells in adrenal lesions derived from FLACs, and if true, possibly also decrease glucocorticoid production from affected glands. Our data support this hypothesis, as both in vitro and in vivo celecoxib decreased steroidogenesis and reversed, at least in part, BAH in the treated animals.
Materials and methods
Animal protocol
AdKO mice have been described previously (Sahut-Barnola et al. 2010). AdKO mice were maintained on a mixed genetic background (C57BL/6/129Sv). There were two groups used in the 6-month-long experiment: celecoxib-treated and vehicle. Celecoxib-treated mice (all female, n=5) were treated with celecoxib 1500 p.p.m. mixed in NIH-31 diet; the vehicle (control) group of AdKO mice (all female, n=4) were on the same NIH-31 diet. Treatment started at the age of 3 months and ended at 9 months of age, because this is the time that these animals develop clinically evident BAH (Sahut-Barnola et al. 2010, de joussineau et al. 2012). We also used only female mice because the BAH phenotype is mostly apparent in them, whereas it is not as penetrant in male mice. Following treatment, mice were euthanized and adrenals harvested for histopathologic and molecular studies. All analyses were done on age-matched mice; animals were from the same litters to avoid strain effects.
All animal studies were performed under protocol ASP12-033 and were approved by and conducted in accordance with the Eunice Kennedy Shriver National Institute for Child Health and Human Development (NICHD) Institutional Animal Care and Use Committee.
Genotyping
PCR was used for genotyping. Three primers (5′-AGCTAGCTTGGCTGGACGTA-3′, 5′-AAGCAGGCGAGCTATTAGTTTAT-3′, and 5′-CATCCATCTCCTATCCCCTTT-3′) were used for Prkar1a genotyping; the WT allele generates a 250 bp fragment while the Prkar1aloxP allele generates a 320 bp product.
Cell cultures and cell counting
Human adrenal cell lines H295R and CAR47 (Nesterova et al. 2008), and the embryonic kidney cell line HEK293 were cultured as monolayers on pretreated T75 flasks. CAR47 was cultured, as previously described (Nesterova et al. 2008) in DMEM (Life Technologies) with 20% fetal bovine serum (FBS) (Gemini Bio-products, West Sacramento, CA, USA) and 1% anti-anti (Life Technologies). H295R was cultured in DMEM/F12 (Life Technologies) with 10% FBS (Gemini Bio-products), 2% NuSerum Culture Supplement (BD Biosciences, San Jose, CA, USA), 1% ITS+Premix Universal Culture Supplement (BD Biosciences), and 1% anti-anti (Life Technologies), while HEK293 was cultured in DMEM with 10% FBS and 1% anti-anti (Life Technologies). Cell numbers were counted using the Luna Automated Cell Counter (Logo Biosystems, Annandale, VA, USA) following Trypan blue staining.
MTT cell viability assay
The Vybrant MTT Cell Proliferation Assay Kit (Life Technologies) was used to examine the effect of celecoxib on cell growth rates in each of the three cell lines: CAR47, H295R, and HEK293. Cells were seeded at 5000 cells/well in 96-well plate and cultured for 48 h. Cells were treated with different concentrations of celecoxib: 1–10 μM diluted in DMSO for 24 h. Cells without treatment were used as control whereas cells treated with DMSO were used as vehicle control. Cell numbers after celecoxib treatment were determined by MTT assay as a representative of cell growth rates. For each well, 10 μl MTT stock solution was added to 100 μl fresh medium and incubated at 37 °C for 4 h. Then 100 μl of SDS–HCl solution was added to each well and incubated at 37 °C for 16 h in a humidified chamber. Absorbance was read at 570 nm.
Cortisol measurements
DetectX Cortisol Enzyme Immunoassay Kit (Arbor Assays, Ann Arbor, MI, USA) was used to exam cortisol concentrations in H295R cell culture medium with and without celecoxib treatment. In each well of a 96-well plate, 5000 cells were seeded. Cells were treated with different concentrations of celecoxib: 1–10 μM diluted in DMSO for 24 h. Cells without treatment were used as control and cells treated with DMSO were used as vehicles. After 24 h, cell culture medium was collected and used to measure cortisol concentrations. Briefly, and following manufacturer's instructions, 50 μl of samples or standards were pipetted into wells in the plate, followed by 25 μl of DetectX cortisol conjugate and 25 μl DetectX cortisol antibody. The plate was incubated at room temperature for 1 h, then washed four times with a wash buffer. After 100 μl of TMB substrate was added to each well and incubated at room temperature for 30 min. Next, 50 μl of the stop solution was added to each well. Optical density generated from each well was read in a plate reader at 450 nm. Cortisol concentrations in cell culture medium with and without celecoxib treatment were calculated based on absorbance.
Haematoxylin and eosin and other immunohistochemistry
Adrenal glands that were dissected from mice were fixed in 10% neutral buffered formalin, embedded into paraffin blocks and sectioned (5 μm) and onto glass slides, following standard procedures by Histoserv, Inc. (Germantown, MD, USA). Histological staining using haematoxylin and eosin staining was also performed (Histoserv, Inc.). Immunohistochemistry (IHC) was performed in house for caspase-3 (rabbit-anti-cleaved-caspase-3 antibody, 1:400, Cell Signaling Technology, Danvers, MA, USA) using the Rabbit ImmPRESS Kit (Vector Laboratories, Burlingame, CA, USA), following manufacturers protocol. Staining was visualized using ImmPACT DAB Peroxidase (HRP) Substrate (Vector Laboratories). Coverslips were mounted onto the stained sections using VectaMount AQ Aqueous Mounting Media (Vector Laboratories). Slides were examined using the Leica DMRX microscope. Images were acquired with the DP72 camera (Olympus) using the CellSans Software (Olympus).
Corticosterone measurements
A mouse corticosterone ELISA (ALPCO Diagnostics, Salem, NH, USA) was used to measure serum corticosterone levels in age-matched 9-month-old celecoxib-treated and control AdKO mice, as per the manufacturer's instructions.
Studies of the glucocorticoid receptor in mouse adrenal glands
RNA studies
Frozen adrenals were disrupted in lysis buffer and total mRNAs were extracted using an RNA extraction kit (Macherey–Nagel, Bethlehem, PA, USA) according to manufacturer's instructions. Total RNAs (1 μg) were reverse-transcribed by Moloney murine leukaemia virus reverse transcriptase (Promega Corp.) according to the manufacturer's instructions. Semi quantitative PCR was stopped at the exponential phase. Primers that were used in this study: glucocorticoid receptor (GR), 5′-CCCTTCCCTTCCCAGATTAG-3′ and 5′-CAGCTCTTTCAGGACCTTCG-3′, 599 bp; β-actin, 5′-CGTGGGCCGCCCTAGGCACCA-3′ and 5′-TTGGCCTTAGGGTTCAGGGGGG-3′, 242 bp.
Protein studies: western blot and immunostaining
Protein was extracted by disruption offrozen adrenals in RIPA buffer in the presence of protease inhibitors (complete mix, Roche). Proteins (30 μg) were loaded on a 8% SDS–PAGE gel, transferred onto nitrocellulose and detected with antibody to GR (1:200, GR(M-20) SC-1004, Santa Cruz Biotechnology). Expression of these proteins was normalized to expression of β-actin and signals were analyzed with Quantity One (Bio-Rad).
Paraffin-embedded AdKO adrenal sections underwent deparaffinization by Histoclear and a graded series of ethanol washes. Followed by immunodetection of GR (1:2000, GR(M-20) SC-1004, Santa Cruz Biotechnology) was carried out on an Intavis InSitu Pro VSi automated processor (Intavis AG, Cologne, Germany). Epitopes were retrieved by 20 min boiling in sodium citrate 10 mM pH 6.0/0.05% Tween-20. Endogenous peroxidases were inactivated with 0.3% hydrogen peroxide during 30 min. Primary antibodies were detected with the appropriate secondary antibodies, coupled to biotin (1:500, Jackson Immunoresearch, West Grove, PA, USA). Biotin was then complexed with streptavidin coupled to HRP (016-030-084, Jackson Immunoresearch). HRP activity was detected with the chromogenic substrate Novared (SK4800, Vector Laboratories, Burlingame, CA, USA). Sections were then mounted in PBS–glycerol and photographed on an Axiophot microscope (Carl Zeiss, Zurich, Switzerland).
PGE2 measurements
A PGE2 ELISA Kit (Abcam, Cambridge, MA, USA) was used to measure serum PGE2 levels in age-matched (9-month-old) control AdKO and celecoxib-treated AdKO mice, as per the manufacturer's instructions.
Ki67 immunofluorescence
Paraffin-embedded sections underwent deparafinization by Histoclear and a graded series of ethanol washes (100, 95, 75, and 50%), followed by antigen retrieval in Antigen Unmasking Solution, Citric Acid Based (Vector Laboratories), in a steamer for 20 min, followed by blocking of non-specific binding using 10% normal goat serum. Slides were incubated with rabbit-anti-ki67 (1:100; Thermo Fisher Scientific, Waltham, MA, USA) primary antibody at 4 °C overnight. Slides were washed three times with 0.1 PBS+0.01% Tween-20 buffer, then incubated with fluorochrome conjugated Alexa Fluor-555 goat anti-rabbit secondary antibody (1:400; Life Technologies) at room temperature for 1 h. Slides were mounted with cover slips using ProLong Gold Antifade Mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Life Technologies) mounting solution. Slides were examined for Ki67 staining using the Leica DMRX microscope equipped with fluorescence capabilities. Images were acquired with the DP72 camera (Olympus) using the CellSans Software (Olympus).
TUNEL assay
Cell apoptotic activities were studied by DeadEnd Fluorometric TUNEL System (Promega). We used paraffin embedded adrenal tissue sections and followed the manufacturer's protocol. Adrenals and staining were visualized using the Leica DMRX fluorescence microscope.
Quantitative real-time PCR
For each sample, 1 μg of mRNA was reverse transcribed to cDNA using High Capacity RNA-to-cDNA Kit (Applied Biosystems). Quantitative real-time PCR was performed in an Applied Biosystems 7500 Real-Time PCR System. mRNA expression levels were detected by caspase-1 (Casp1, Mm00438023_m1), interleukin 1 beta (Il1b, Mm00434228_m1), PTGS2 (COX2) (Ptgs2, Mm00478374_m1), sf1 (Nr5a1, Mm00446826_m1), β-catenin (Ctnnb1, Mm00483039_m1), cyclin D1 (Ccnd1, Mm00432359_m1), and cyp17a (Cyp17a1, Mm00484040_m1) TaqMan Gene Expression Assays (Applied Biosystems). Gene expression levels were normalized to Gapdh (TaqMan Rodent Gapdh control, Applied Biosystems). Gene expression in celecoxib-treated mouse adrenal glands was calculated relative to control AdKO adrenal glands using 2−ΔΔCt values.
Statistical analysis
All analyses were performed using the two-tailed Student's t-test. Differences at P<0.05 were considered statistically significant. Mean±s.e.m. of at least in-triplicate experiments were calculated. Animals were age- and sex-matched for all comparisons between treated and untreated.
Results
Responses of human cell lines to celecoxib
Human adrenocortical cell lines CAR47 and H295R were treated with different concentrations of celecoxib. CAR47 is the first (and only) immortalized human cell line with an inactivating PRKAR1A mutation. It was derived from the adrenal gland of a patient diagnosed with PPNAD and CNC (Nesterova et al. 2008). H295R is an adrenocortical carcinoma cell line that retains the ability to produce adrenal steroids (Rainey et al. 1994), including cortisol, whereas CAR47 does not (Rainey et al. 1994, Wang et al. 2012). HEK293 is a human embryonic kidney cell line (Graham et al. 1977) and was used here as a ‘control’ line against the immortalized adrenal cells of H295R and CAR47.
The three cell lines were treated with different concentrations of celecoxib for 24 h. Cell viabilities were then studied by MTT cell proliferation assay and the cell number was registered. Increased celecoxib concentrations led to decreasing cell viability. The CAR47 and H295R lines showed similar responses that were significantly different from HEK293 at 1, 3, 4, and 7 μM concentrations of celecoxib (Fig. 1A). Increasing celecoxib concentrations led to decreasing cell numbers with the CAR47 and H295R lines with fewer cells compared to HEK293 (Fig. 1B). In addition, celecoxib concentrations more than 2 μM reduced cortisol concentrations in H295R cell culture medium (Fig. 2A). However, if we normalize the decrease in cortisol concentrations by the number of cells, it appears that the decrease in cortisol levels was due to the effect of celecoxib on cell viability (Fig. 2B).
Effect of celecoxib treatment on human adrenal cell lines CAR47 and H295R, with HER293 cell line used as control. (A) Cell viability studied by the MTT cell proliferation assay: decreased cell viability was found in the CAR47 and H295R cell lines with increasing celecoxib concentrations. (B) Cell number changes by cell counting revealed decreasing cell numbers with increasing celecoxib concentrations. The decrease in the two human adrenal cell lines became more significant at higher celecoxib concentrations. *, P<0.05.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
(A) Effect of celecoxib treatment on cortisol concentrations in the human adrenal cell line H295R. Increasing celecoxib concentrations led to decreasing cortisol levels in cell culture medium. This effect became statistically significant at concentrations equal to or higher than 2 μM. (B) The decrease in cortisol levels in cell culture medium is mainly due to the effect of celecoxib on cell viability. *, P<0.05.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Celecoxib treatment of AdKO mice: effects on the adrenal and corticosterone levels
Following 6 months of celecoxib treatment, partial rescue of the phenotype caused by adrenocortical Prkar1a loss (Sahut-Barnola et al. 2010, de Joussineau et al. 2012) was observed: BAH of the adrenal cortex was seen in all 9-month-old AdKO mice that did not receive celecoxib, whereas none of the 9-month-old treated mice had clear evidence of BAH (Fig. 3). Moreover, plasma corticosterone levels in celecoxib-treated AdKO mice were significantly reduced compared to those in 9-month-old age-matched vehicle-treated AdKO mice after the 6-month-long treatment (Fig. 4A). This was accompanied by a decrease in serum PGE2 levels in the age-matched AdKO mice following the 6-month-long treatment (Fig. 4B).
Celecoxib treatment of AdKO female mice was associated with lack of development of adrenocortical hyperplasia. Hematoxylin and eosin (H&E) staining from adrenal glands of AdKO mice untreated (control) and treated (celecoxib). All female untreated AdKO mice developed hyperplasia by 9 months of age, while no obvious hyperplastic areas were seen in any of the treated animals.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Changes in levels of corticosterone and PGE2 after celecoxib treatment: (A) corticosterone: 6 months of celecoxib treatment (between the ages of 3 and 9 months) led to reduced corticosterone levels. (B) PGE2: likewise, celecoxib treatment led to decreased PGE2 level between the beginning of the treatment (3 months of age) and the end (9 months of age). *, P<0.05.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Ki67 immunofluorescence was used for the assessment of cell proliferation of AdKO mouse adrenals (Fig. 5A), as previously described (Schluter et al. 1993). The number of Ki67-positive cells was reduced in the adrenal glands of celecoxib-treated mice compared to vehicle (Fig. 5B).
Effect of celecoxib on adrenocortical cell proliferation: (A) Ki67 staining in adrenals from mice untreated (control) and treated (celecoxib). Ki67 positive staining (red) and nuclear staining (blue). (B) As shown by quantification of Ki67-positive cells, 6 months of celecoxib treatment led to decreased proliferation in the adrenal cortex of AdKO mice.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Celecoxib treatment rescued the resistance of AdKO adrenocortical cells to apoptosis, as measured by caspase-3 IHC (Fig. 6A) and TUNEL fluorescent assay (Fig. 6B), which were performed as previously described (Labat-Moleur et al. 1998). Apoptotic cells were detected only in the adrenal gland of celecoxib-treated mice, but not in those of untreated animals by both IHC and fluorescence (Fig. 6A and B).
Celecoxib treatment triggered apoptotic activity in the adrenal cortex of AdKO mice: (A) caspase-3 staining: no staining seen in the adrenal cortex of AdKO mice that received no treatment at the age of 9 months, whereas in mice treated with celecoxib, there were several caspase-3-positive cells. (B) TUNEL assay. Likewise, TUNEL-positive cells were seen (red arrows) only in the adrenal glands of celecoxib-treated animals, whereas no TUNEL-positive cells were seen in AdKO mice that were not treated with celecoxib.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Changes in gene expression after celecoxib treatment
RT-PCR analysis showed significant effects of 6 months of dietary celecoxib treatment on a number of genes. We found decreased expression of Casp1, Il1b, COX2 (Ptgs2), SF1 (Nr5a1), β-catenin (Ctnnb1), and cyclin D1 (Ccnd1) in the adrenals of celecoxib-treated mice (Fig. 7A). We also analyzed the expression of different enzymes involved in the steroidogenesis. We found decreased gene expression of cyp11a and cyp17a, two enzymes involved in steroidogenesis in celecoxib-treated AdKO mice (Fig. 7B) compared with age-matched AdKO control, although only cyp17a expression was statistically significantly reduced. Thus, celecoxib treatment corrects at least in part the overexpression of the Cyp17a enzyme that was previously observed (Sahut-Barnola et al. 2010) in AdKO mice.
(A) Real-time PCR analysis of genes related to inflammation and steroidogenesis: celecoxib treatment led to significant reductions in caspase-1, interleukin-1b, and COX2 (Ptgs2), steroidogenic factor 1, β-catenin, and cell cycle regulator cyclin D1. (B) Real-time PCR analysis of genes related to steroidogenesis. Celecoxib treatment decreased cyp11a and cyp17a expression, although only the latter was significant.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Investigation of the GR in treated and untreated mice
Celecoxib has been shown to enhance the function of the GR (Hu et al. 2005) and GR has been implicated in human (Bourdeau et al. 2003) but not mouse (Louiset et al. 2009) BAH. We studied GR expression at the mRNA and protein level in AdKO mice both at baseline and after treatment with celecoxib (Fig. 8). AdKO mice showed no paradoxical increase in glucocorticoid secretion (data not shown) and there were no differences at the mRNA or at the protein levels by immunoblotting at baseline (Fig. 8A and B). Immunolocalization showed that GR was not even present in AdKO cortical steroidogenic cells and was mainly detected in non-steroidogenic cells (capsule, blood vessels and, as expected, in chromaffin cells; Fig. 8C). When we compared total GR expression in celecoxib-treated and vehicle, there were also no differences (Fig. 8D).
(A) Representative agarose gel electrophoresis of GR cDNA amplified (RT-PCR) from adrenal mRNA of WT and AdKO 12-month-old female mice. (B) Western blot showing GR protein signal in WT and AdKO adrenal extracts from 12-month-old female mice. Antibody to glucocorticoid receptor: GR(M-20) SC-1004 (Santa Cruz Biotechnology). (C) GR immunostaining in adrenal sections from 8-month-old WT/AdKO females. GR staining is detected (arrows) in chromaffin cells nuclei, in capsular cells (top magnification) and in endothelial cells (bottom magnification) and is barely detectable in steroidogenic cells. M, medulla; scale bars, 20 μm. (D) Immunofluorescence staining of GR in 9-month-old vehicle-treated AdKO mice and 9-month-old celecoxib-treated AdKO mice: there are no differences in GR expression between treated and untreated animals.
Citation: Endocrine-Related Cancer 23, 1; 10.1530/ERC-15-0472
Discussion
Celecoxib, uniquely among all non-steroid anti-inflammatory drugs (NSAIDs), has been used extensively both in vitro and in vivo to inhibit neoplastic growth or proliferation (Bertagnolli 2006, Daikoku et al. 2014, Yokouchi et al. 2014, Kurtova et al. 2015). Some of the in vitro effects may be because at certain concentrations the drug is insoluble and, thus, causes precipitate-dependent effects and cellular toxicity (Sacchetti 2013). However, there are also true biologic effects of celecoxib, mostly linked to PGE2 inhibition and the consequent decrease of this molecule's induction of stem cell-like cells that are essential for the growth of at least some tumors (Kurtova et al. 2015).
We hypothesized that celecoxib might be useful in the treatment of PPNAD, because in both animals (Sahut-Barnola et al. 2010, de Joussineau et al. 2012) and humans with various forms of BAH (Carney et al. 2010, 2011, 2012, Almeida & Stratakis 2011), the disease has been linked to FLACs. In addition, in at least two other lesions caused by Prkar1a deficiency, mouse osteochondromyxomas and FDL tumors, PGE2 levels were elevated both at the tumor level and systemically, as well as linked to abnormal growth and proliferation of aBSCs, relatively undifferentiated cells of the osteoblastic lineage (Almeida et al. 2011). Indeed, our data support an effect of systemic celecoxib therapy in mouse BAH caused by Prkar1a deficiency, in addition to an inhibitory effect on human adrenal cell viability in culture.
COX2 (PTGS2) is expressed in the human and rodent adrenal cortex and cell lines, as previous studies have demonstrated (Salmenkivi et al. 2001, Engstrom et al. 2008, Martinez Calejman et al. 2011). Additionally, COX2 is activated in cells of the immune and/or inflammation system embedded within the cortex (Cover et al. 2001, Ichitani et al. 2001). In turn, corticosteroids suppress COX2 expression (Zhang et al. 1999). The present study suggested that celecoxib inhibition of COX2 led to decreased PGE2 levels and glucocorticoid secretion, but more importantly an improvement in the histologic phenotype of the BAH associated with Prkar1a deficiency. It is tempting to speculate that the decreased proliferation and increased apoptosis within the mouse adrenals affected by BAH and treated with celecoxib were due to an effect on the primary cells responsible for the phenotype: Prkar1a-null cells that are FLACs, whose continued proliferation in adult adrenal glands cause hyperplasia and tumors (Sahut-Barnola et al. 2010, de Joussineau et al. 2012). PGE2 is essential for maintaining stem cell-like cells in a number of tissues (Cha et al. 2006) and inhibiting the enzyme that degrades enhances regenerative capacity (Essex et al. 2013).
Our data support a possible role of celecoxib in the treatment of patients with PPNAD. However, human studies have not been performed. Celecoxib is widely used and there are a number of known side effects, but adrenocortical insufficiency is not one of them (Zhang et al. 2015). The latter may be because our data do not support an effect of celecoxib on normal adrenal cortex, but only in these glands where there is hyperplasia and/or FLACs.
The observed effects were not due to differences in GR expression in celecoxib-treated animals (Fig. 8). Celecoxib has been shown to enhance the function of the GR (Hu et al. 2005) and GR has been implicated in human (Bourdeau et al. 2003, Briassoulis et al. 2011) but not mouse (Louiset et al. 2009) forms of BAH. Humans with PRKAR1A mutations and BAH (mostly PPNAD) demonstrate paradoxical increase in their glucocorticoid secretion in response to dexamethasone (Stratakis et al. 1999). However, this has not been seen in any of the animal models of PPNAD studied to date (Griffin et al. 2004a,b, Louiset et al. 2009, Sahut-Barnola et al. 2010) and it is not universally present among patients with BAH and PRKAR1A mutations (Stratakis & Kirschner 1998, Stratakis et al. 1999, Hofland et al. 2012). GR expression was not increased in AdKO mice, the GR was barely present in adrenocortical steroidogenic cells, and its expression was not modified by celecoxib (Fig. 8).
In conclusion, our study of celecoxib, a COX2 inhibitor, showed promising results in the treatment of mice with a form of BAH caused by Prkar1a deficiency. Follow-up studies are needed but it is probable that NSAIDs that decrease PGE2 levels may provide a useful alternative to steroidogenesis inhibitors for patients with PPNAD.
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 supported by the Intramural Research Program (IRP) of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), National Institutes of Health (NIH), Bethesda, Maryland 20892, USA.
References
Almeida MQ & Stratakis CA 2011 How does cAMP/protein kinase A signaling lead to tumors in the adrenal cortex and other tissues? Molecular and Cellular Endocrinology 336 162–168. (doi:10.1016/j.mce.2010.11.018).
Almeida MQ, Tsang KM, Cheadle C, Watkins T, Grivel JC, Nesterova M, Goldbach-Mansky R & Stratakis CA 2011 Protein kinase A regulates caspase-1 via Ets-1 in bone stromal cell-derived lesions: a link between cyclic AMP and pro-inflammatory pathways in osteoblast progenitors. Human Molecular Genetics 20 165–175. (doi:10.1093/hmg/ddq455).
Bertagnolli MM 2006 Celecoxib for the prevention of sporadic colorectal adenomas. New England Journal of Medicine 355 873–884. (doi:10.1056/NEJMoa061355).
Bertherat J, Horvath A, Groussin L, Grabar S, Boikos S, Cazabat L, Libe R, René-Corail F, Stergiopoulos S & Bourdeau I et al. 2009 Mutations in regulatory subunit type 1A of cyclic adenosine 5′-monophosphate-dependent protein kinase (PRKAR1A): phenotype analysis in 353 patients and 80 different genotypes. Journal of Clinical Endocrinology and Metabolism 94 2085–2089. (doi:10.1210/jc.2008-2333).
Bourdeau I, Lacroix A, Schürch W, Caron P, Antakly T & Stratakis CA 2003 Primary pigmented nodular adrenocortical disease: paradoxical responses of cortisol secretion to dexamethasone occur in vitro and are associated with increased expression of the glucocorticoid receptor. Journal of Clinical Endocrinology and Metabolism 88 3931–3937. (doi:10.1210/jc.2002-022001).
Briassoulis G, Damjanovic S, Xekouki P, Lefebvre H & Stratakis CA 2011 The glucocorticoid receptor and its expression in the anterior pituitary and the adrenal cortex: a source of variation in hypothalamic–pituitary–adrenal axis function; implications for pituitary and adrenal tumors. Endocrine Practice 17 941–948. (doi:10.4158/EP11061.RA).
Carney JA, Gordon H, Carpenter PC, Shenoy BV & Go VL 1985 The complex of myxomas, spotty pigmentation, and endocrine overactivity. Medicine 64 270–283. (doi:10.1097/00005792-198507000-00007).
Carney JA, Boccon-Gibod L, Jarka D, Tanaka Y, Swee RG, Unni KK & Stratakis CA 2001 Osteochondromyxoma of bone: a congenital tumor associated with lentigines and other unusual disorders. American Journal of Surgical Pathology 25 164–176. (doi:10.1097/00000478-200102000-00004).
Carney JA, Gaillard RC, Bertherat J & Stratakis CA 2010 Familial micronodular adrenocortical disease, Cushing syndrome, and mutations of the gene encoding phosphodiesterase 11A4 (PDE11A). American Journal of Surgical Pathology 34 547–555. (doi:10.1097/PAS.0b013e3181d31f49).
Carney JA, Young WF & Stratakis CA 2011 Primary bimorphic adrenocortical disease: cause of hypercortisolism in McCune–Albright syndrome. American Journal of Surgical Pathology 35 1311–1326. (doi:10.1097/PAS.0b013e31821ec4ce).
Carney JA, Ho J, Kitsuda K, Young WF Jr & Stratakis CA 2012 Massive neonatal adrenal enlargement due to cytomegaly, persistence of the transient cortex, and hyperplasia of the permanent cortex: findings in Cushing syndrome associated with hemihypertrophy. American Journal of Surgical Pathology 36 1452–1463. (doi:10.1097/PAS.0b013e31825d538b).
Cha YI, Solnica-Krezel L & DuBois RN 2006 Fishing for prostanoids: deciphering the developmental functions of cyclooxygenase-derived prostaglandins. Developmental Biology 289 263–272. (doi:10.1016/j.ydbio.2005.10.013).
Cover PO, Slater D & Buckingham JC 2001 Expression of cyclooxygenase enzymes in rat hypothalamo-pituitary–adrenal axis: effects of endotoxin and glucocorticoids. Endocrine 16 123–131. (doi:10.1385/ENDO:16:2:123).
Daikoku T, Terakawa J, Hossain MM, Yoshie M, Cappelletti M, Yang P, Ellenson LH & Dey SK 2014 Mammalian target of rapamycin complex 1 and cyclooxygenase 2 pathways cooperatively exacerbate endometrial cancer. American Journal of Surgical Pathology 184 2390–2402. (doi:10.1016/j.ajpath.2014.05.023).
Engstrom L, Rosén K, Angel A, Fyrberg A, Mackerlova L, Konsman JP, Engblom D & Blomqvist A 2008 Systemic immune challenge activates an intrinsically regulated local inflammatory circuit in the adrenal gland. Endocrinology 149 1436–1450. (doi:10.1210/en.2007-1456).
Essex MN, Zhang RY, Berger MF, Upadhyay S & Park PW 2013 Safety of celecoxib compared with placebo and non-selective NSAIDs: cumulative meta-analysis of 89 randomized controlled trials. Expert Opinion on Drug Safety 12 465–477. (doi:10.1517/14740338.2013.780595).
Graham FL, Smiley J, Russell WC & Nairn R 1977 Characteristics of a human cell line transformed by DNA from human adenovirus type 5. Journal of General Virology 36 59–74. (doi:10.1099/0022-1317-36-1-59).
Greenhough A, Smartt HJ, Moore AE, Roberts HR, Williams AC, Paraskeva C & Kaidi A 2009 The COX-2/PGE2 pathway: key roles in the hallmarks of cancer and adaptation to the tumour microenvironment. Carcinogenesis 30 377–386. (doi:10.1093/carcin/bgp014).
Griffin KJ, Kirschner LS, Matyakhina L, Stergiopoulos S, Robinson-White A, Lenherr S, Weinberg FD, Claflin E, Meoli E & Cho-Chung YS et al. 2004a Down-regulation of regulatory subunit type 1A of protein kinase A leads to endocrine and other tumors. Cancer Research 64 8811–8815. (doi:10.1158/0008-5472.CAN-04-3620).
Griffin KJ, Kirschner LS, Matyakhina L, Stergiopoulos SG, Robinson-White A, Lenherr SM, Weinberg FD, Claflin ES, Batista D & Bourdeau I et al. 2004b A transgenic mouse bearing an antisense construct of regulatory subunit type 1A of protein kinase A develops endocrine and other tumours: comparison with Carney complex and other PRKAR1A induced lesions. Journal of Medical Genetics 41 923–931. (doi:10.1136/jmg.2004.028043).
Hofland J, de Herder WW, Derks L, Hofland LJ, van Koetsveld PM, de Krijger RR, van Nederveen FH, Horvath A, Stratakis CA & de Jong FH et al. 2012 Regulation of steroidogenesis in a primary pigmented nodular adrenocortical disease-associated adenoma leading to virilization and subclinical Cushing's syndrome. European Journal of Endocrinology 168 67–74. (doi:10.1530/EJE-12-0594).
Hu F, Wang X, Pace TW, Wu H & Miller AH 2005 Inhibition of COX-2 by celecoxib enhances glucocorticoid receptor function. Molecular Psychiatry 10 426–428. (doi:10.1038/sj.mp.4001644).
Ichitani Y, Holmberg K, Maunsbach AB, Haeggström JZ, Samuelsson B, De Witt D & Hokfelt T 2001 Cyclooxygenase-1 and cyclooxygenase-2 expression in rat kidney and adrenal gland after stimulation with systemic lipopolysaccharide: in situ hybridization and immunocytochemical studies. Cell and Tissue Research 303 235–252. (doi:10.1007/s004410000296).
de Joussineau C, Sahut-Barnola I, Levy I, Saloustros E, Val P, Stratakis CA & Martinez A 2012 The cAMP pathway and the control of adrenocortical development and growth. Molecular and Cellular Endocrinology 351 28–36. (doi:10.1016/j.mce.2011.10.006).
Kirschner LS, Carney JA, Pack SD, Taymans SE, Giatzakis C, Cho YS, Cho-Chung YS & Stratakis CA 2000 Mutations of the gene encoding the protein kinase A type I-α regulatory subunit in patients with the Carney complex. Nature Genetics 26 89–92. (doi:10.1038/79238).
Kirschner LS, Kusewitt DF, Matyakhina L, Towns WH, Carney JA, Westphal H & Stratakis CA 2005 A mouse model for the Carney complex tumor syndrome develops neoplasia in cyclic AMP-responsive tissues. Cancer Research 65 4506–4514. (doi:10.1158/0008-5472.CAN-05-0580).
Kurtova AV, Xiao J, Mo Q, Pazhanisamy S, Krasnow R, Lerner SP, Chen F, Roh TT, Lay E & Ho PL et al. 2015 Blocking PGE2-induced tumour repopulation abrogates bladder cancer chemoresistance. Nature 517 209–213. (doi:10.1038/nature14034).
Labat-Moleur F, Guillermet C, Lorimier P, Robert C, Lantuejoul S, Brambilla E & Negoescu A 1998 TUNEL apoptotic cell detection in tissue sections: critical evaluation and improvement. Journal of Histochemistry and Cytochemistry 46 327–334. (doi:10.1177/002215549804600306).
Louiset E, Stratakis CA, Perraudin V, Griffin KJ, Libé R, Cabrol S, Fève B, Young J, Groussin L & Bertherat J et al. 2009 The paradoxical increase in cortisol secretion induced by dexamethasone in primary pigmented nodular adrenocortical disease involves a glucocorticoid receptor-mediated effect of dexamethasone on protein kinase A catalytic subunits. Journal of Clinical Endocrinology and Metabolism 94 2406–2413. (doi:10.1210/jc.2009-0031).
Martinez Calejman C, Astort F, Di Gruccio JM, Repetto EM, Mercau M, Giordanino E, Sanchez R, Pignataro O, Arias P & Cymeryng CB 2011 Lipopolysaccharide stimulates adrenal steroidogenesis in rodent cells by a NFκB-dependent mechanism involving COX-2 activation. Molecular and Cellular Endocrinology 337 1–6. (doi:10.1016/j.mce.2010.12.036).
Nesterova M, Bossis I, Wen F, Horvath A, Matyakhina L & Stratakis CA 2008 An immortalized human cell line bearing a PRKAR1A-inactivating mutation: effects of overexpression of the wild-type allele and other protein kinase A subunits. Journal of Clinical Endocrinology and Metabolism 93 565–571. (doi:10.1210/jc.2007-1902).
Rainey WE, Bird IM & Mason JI 1994 The NCI-H295 cell line: a pluripotent model for human adrenocortical studies. Molecular and Cellular Endocrinology 100 45–50. (doi:10.1016/0303-7207(94)90277-1).
Sacchetti A 2013 Cancer cell killing by celecoxib: reality or just in vitro precipitation-related artifact? Journal of Cellular Biochemistry 114 1434–1444. (doi:10.1002/jcb.24485).
Sahut-Barnola I, de Joussineau C, Val P, Lambert-Langlais S, Damon C, Lefrancois-Martinez AM, Pointud JC, Marceau G, Sapin V & Tissier F 2010 Cushing's syndrome and fetal features resurgence in adrenal cortex-specific Prkar1a knockout mice. PLoS Genetics 6 e1000980. (doi:10.1371/journal.pgen.1000980).
Salmenkivi K, Haglund C, Ristimaki A, Arola J & Heikkila P 2001 Increased expression of cyclooxygenase-2 in malignant pheochromocytomas. Journal of Clinical Endocrinology and Metabolism 86 5615. (doi:10.1210/jcem.86.11.8052).
Schluter C, Duchrow M, Wohlenberg C, Becker MH, Key G, Flad HD & Gerdes J 1993 The cell proliferation-associated antigen of antibody Ki-67: a very large, ubiquitous nuclear protein with numerous repeated elements, representing a new kind of cell cycle-maintaining proteins. Journal of Cell Biology 123 513–522. (doi:10.1083/jcb.123.3.513).
Stratakis CA 2007 Adrenocortical tumors, primary pigmented adrenocortical disease (PPNAD)/Carney complex, and other bilateral hyperplasias: the NIH studies. Hormone and Metabolic Research 39 467–473. (doi:10.1055/s-2007-981477).
Stratakis CA 2008 Cushing syndrome caused by adrenocortical tumors and hyperplasias (corticotropin-independent Cushing syndrome). Endocrine Development 13 117–132.
Stratakis CA & Kirschner LS 1998 Clinical and genetic analysis of primary bilateral adrenal diseases (micro- and macronodular disease) leading to Cushing syndrome. Hormone and Metabolic Research 30 456–463. (doi:10.1055/s-2007-978914).
Stratakis CA, Sarlis N, Kirschner LS, Carney JA, Doppman JL, Nieman LK, Chrousos GP & Papanicolaou DA 1999 Paradoxical response to dexamethasone in the diagnosis of primary pigmented nodular adrenocortical disease. Annals of Internal Medicine 131 585–591. (doi:10.7326/0003-4819-131-8-199910190-00006).
Tsang KM, Starost MF, Nesterova M, Boikos SA, Watkins T, Almeida MQ, Harran M, Li A, Collins MT & Cheadle C et al. 2010 Alternate protein kinase A activity identifies a unique population of stromal cells in adult bone. PNAS 107 8683–8688. (doi:10.1073/pnas.1003680107).
Wang T, Rowland JG, Parmar J, Nesterova M, Seki T & Rainey WE 2012 Comparison of aldosterone production among human adrenocortical cell lines. Hormone and Metabolic Research 44 245–250. (doi:10.1055/s-0031-1298019).
Yokouchi H, Kanazawa K, Ishida T, Oizumi S, Shinagawa N, Sukoh N, Harada M, Ogura S, Munakata M & Dosaka-Akita H 2014 Cyclooxygenase-2 inhibitors for non-small-cell lung cancer: a phase II trial and literature review. Molecular and Clinical Oncology 2 744–750.
Zhang MZ, Harris RC & McKanna JA 1999 Regulation of cyclooxygenase-2 (COX-2) in rat renal cortex by adrenal glucocorticoids and mineralocorticoids. PNAS 96 15280–15285. (doi:10.1073/pnas.96.26.15280).
Zhang Y, Desai A, Yang SY, Bae KB, Antczak MI, Fink SP, Tiwari S, Willis JE, Williams NS & Dawson DM et al. 2015 Inhibition of the prostaglandin-degrading enzyme 15-PGDH potentiates tissue regeneration. Science 348 6240.
