novel genes involved in pathophysiology of gonadotropin dependent adrenal tumors in mice

Accepted Manuscript Novel genes involved in pathophysiology of gonadotropin-dependent adrenal tumors in mice Milena Doroszko, Marcin Chrusciel, Kirstine Belling, Susanna Vuorenoja, Marlene Dalgaard, Henrik Leffers, H Bjørn Nielsen, Ilpo Huhtaniemi, Jorma Toppari, Nafis A Rahman PII: S0303-7207(17)30049-7 DOI: 10.1016/j.mce.2017.01.036 Reference: MCE 9809 To appear in: Molecular and Cellular Endocrinology Received Date: December 2016 Revised Date: 21 January 2017 Accepted Date: 22 January 2017 Please cite this article as: Doroszko, M., Chrusciel, M., Belling, K., Vuorenoja, S., Dalgaard, M., Leffers, H., Nielsen, H.B., Huhtaniemi, I., Toppari, J., Rahman, N.A., Novel genes involved in pathophysiology of gonadotropin-dependent adrenal tumors in mice, Molecular and Cellular Endocrinology (2017), doi: 10.1016/j.mce.2017.01.036 This is a PDF file of an unedited manuscript that has been accepted for publication As a service to our customers we are providing this early version of the manuscript The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain ACCEPTED MANUSCRIPT Novel genes involved in pathophysiology of gonadotropin-dependent adrenal tumors in mice Milena Doroszko1, Marcin Chrusciel1, Kirstine Belling2, Susanna Vuorenoja1, Marlene Dalgaard2, Henrik Leffers2, H Bjørn Nielsen2, Ilpo Huhtaniemi1,3, Jorma Toppari1,4 and Nafis A Rahman1,5 RI PT Department of Physiology, Institute of Biomedicine, University of Turku, Finland Center for Biological Sequence Analysis, Technical University of Denmark, Lyngby, Denmark Institute of Reproductive and Developmental Biology, Imperial College London, London, U.K 10 Department of Pediatrics, Turku University Hospital, Turku, Finland 11 Department of Reproduction and Gynecological Endocrinology, Medical University of Bialystok, 12 Poland M AN U SC 13 15 TE D 14 Word count: 4201 16 Corresponding author and person to whom reprint request should be addressed: 18 Nafis Rahman, MD, PhD 19 University of Turku 20 Institute of Biomedicine, Department Physiology, 21 Kiinamyllynkatu 10, Turku 20520, FINLAND 22 e-mail: nafis.rahman@utu.fi AC C EP 17 23 24 25 26 ACCEPTED MANUSCRIPT ABSTRACT 28 Specific inbred strains and transgenic inhibin-α Simian Virus 40 T antigen (inhα/Tag) mice are 29 genetically susceptible to gonadectomy-induced adrenocortical neoplasias We identified altered gene 30 expression in prepubertally gonadectomized (GDX) inhα/Tag and wild-type (WT) mice Besides 31 earlier reported Gata4 and Lhcgr, we found up-regulated Esr1, Prlr-rs1, and down-regulated Grb10, 32 Mmp24, Sgcd, Rerg, Gnas, Nfatc2, Gnrhr, Igf2 in inhα/Tag adrenal tumors Sex-steroidogenic enzyme 33 genes expression (Srd5a1, Cyp19a1) was up-regulated in tumors, but adrenal-specific steroidogenic 34 enzyme (Cyp21a1, Cyp11b1, Cyp11b2) down-regulated We localized novel Lhcgr transcripts in 35 adrenal cortex parenchyma and in non-steroidogenic A cells, in GDX WT and in intact WT mice We 36 identified up-regulated Esr1 as a potential novel biomarker of gonadectomy-induced adrenocortical 37 tumors in inhα/Tag mice presenting with an inverted adrenal-to-gonadal steroidogenic gene expression 38 profile A putative normal adrenal remodeling or tumor suppressor role of the down-regulated genes 39 (e.g Grb10, Rerg, Gnas, and Nfatc2) in the tumors remains to be addressed M AN U SC RI PT 27 TE D 40 41 EP 42 43 AC C 44 45 46 47 48 49 Keywords: adrenal tumor, pathophysiology, gonadotropin, GATA4, LHCGR, biomarker ACCEPTED MANUSCRIPT 50 INTRODUCTION The adrenal cortex is the steroidogenically active part of the adrenal gland consisting of 52 layers, i.e., zona glomerulosa, zona fasciculata and zona reticularis in humans, or x-zone in mice 53 (corresponding to human zona reticularis), which is a vestige of the fetal adrenal cortex (Pihlajoki, et 54 al 2015) Each layer expresses specific steroidogenic enzymes that results in the zone-specific 55 production of mineralocorticoids, glucocorticoids or androgens, respectively (Pihlajoki et al 2015) In 56 contrast to human zona reticularis, the x-zone in postnatal mice does not produce androgens due to 57 inactivation by methylation of Cyp17a1, the enzyme directing steroidogenesis to production of sex 58 steroids (Pihlajoki et al 2015) The highly differentiated cells of the adrenal zones not proliferate, 59 but are replaced by new cells migrating towards medulla from the subcapsular niche of 60 stem/progenitor cells (Pihlajoki et al 2015) This pool of progenitors has a common origin with 61 gonadal somatic cells, which forms the adreno-gonadal primordium in early embryonic life (Laufer, et 62 al 2012; Pihlajoki et al 2015) Interestingly, following gonadectomy, some strains of inbred mice 63 (DBA/2J, NU/J) develop adrenocortical neoplasms that structurally and functionally resemble ovarian 64 theca and granulosa cells (Bielinska, et al 2005; Bielinska, et al 2003) TE D M AN U SC RI PT 51 Prepubertally gonadectomized inbred DBA/2J or NU/J mice develop adrenal adenomas with 66 two types of neoplastic cells, non-steroidogenic A cells (gonadectomy-independent) originating from 67 the subcapsular region (Hofmann, et al 1960) and larger lipid-laden B cells (gonadectomy-dependent) 68 located between the foci of A cells (Rosner, et al 1966), but still with unclear clonal origin Only the 69 B cells are steroidogenic and express gonad-specific genes such as Lhcgr, Gata4, and the 70 steroidogenic enzymes Cyp17a1 and Cyp19a1, resulting in a significant estrogen production 71 (Bielinska et al 2003; Chrusciel, et al 2013; Krachulec, et al 2012) However, unlike the 72 aforementioned mouse models, human adrenocortical tumors (ACT) produce mostly adrenal steroids 73 (Stojadinovic, et al 2003) and sex steroid production occurs only in 5% of all ACT (Moreno, et al 74 2006) Interestingly, a large group of ACTs respond to luteinizing hormone (LH)/choriogonadotropin 75 (CG) stimuli (Carlson 2007) with production of cortisol, aldosterone or androgens This situation is AC C EP 65 ACCEPTED MANUSCRIPT 76 usually correlated with chronically elevated LH (menopause) or hCG (pregnancy) levels (Alevizaki, et 77 al 2006; Carlson 2007; Saner-Amigh, et al 2006) A common cause for human adrenal malignancies are TP53 germline (80% of pediatric and 79 6% of adult ACT cases) and somatic (25–70% of adult ACC) mutations (Lerario, et al 2014) 80 Inactivation of p53 promotes development of neoplastic cells by enabling progression through the S- 81 phase of the cell cycle, despite DNA damage (Levine, et al 1991) Modeling of p53 inhibition is 82 possible by transgenic expression of the Simian Virus 40 T-antigen (SV40Tag), and to date it has been 83 successfully used to induce malignancies in various mouse organs (Hudson and Colvin 2016) 84 Identifying the molecular mechanisms leading to tumor formation and progression in animal models 85 expressing SV40Tag might therefore provide novel insights into human tumors induced by p53 86 inactivation M AN U SC RI PT 78 Inhα/Tag mice express SV40Tag oncogene under the inhibin α promoter in a C57Bl/6 genetic 88 background and have been well established as a model for gonadal and adrenocortical tumorigenesis 89 (Kananen, et al 1996; Rahman, et al 2004; Rilianawati, et al 1998) Intact inhα/Tag mice have been 90 extensively used to study gonadal tumorigenesis (Kananen et al 1996; Kiiveri, et al 1999; Rahman et 91 al 2004; Vuorenoja, et al 2007) To induce adrenocortical tumorigenesis, the inhα/Tag mice need to 92 be prepubertally gonadectomized (Kananen et al 1996) Generally, females develop adrenal tumors 93 faster and to larger size than males, with high individual variations of tumor volume and histological 94 appearance (Rahman et al 2004) The tumors express highly inhibin alpha (INHA), transcription 95 factor GATA4 and luteinizing hormone/chorionic gonadotropin receptor (LHCGR) (Kananen et al 96 1996; Kiiveri et al 1999; Rahman et al 2004; Rilianawati et al 1998) Adrenocortical tumorigenesis 97 in inhα/Tag model has been shown to be strictly gonadotropin dependent Crossbreeding with mice 98 lacking gonadotropin releasing hormone (GnRH) (hpg mice) or treatment with the GnRH antagonist 99 cetrorelix prevented tumor formation (Rilianawati, et al 2000) Moreover, crossbreeding the inhα/Tag 100 mice with transgenic mice expressing bovine Lhb fused with the human chorionic gonadotropin β- 101 subunit C-terminal peptide (bLHβ-CTP) (Risma, et al 1995), resulted in a 10-fold higher LH levels, 102 and enhanced simultaneous development of gonadal and adrenocortical tumors (Mikola, et al 2003) AC C EP TE D 87 ACCEPTED MANUSCRIPT 103 Until now, molecular mechanisms of the adrenocortical tumor formation and progression in 104 conjunction with the elevated LH and subsequent LHCGR up-regulation remain largely unknown In this study, we used the inhα/Tag male mouse model to analyze the molecular basis of the 106 LH-dependent adrenal tumorigenesis In doing so, we identified and validated novel biomarker genes, 107 complementing the earlier established LHCGR and GATA4 markers We also revisited the Lhcgr 108 transcript localization in normal and neoplastic adrenal cells 109 110 SC 111 RI PT 105 112 M AN U 113 114 115 116 TE D 117 118 119 EP 120 121 AC C 122 123 124 125 126 127 128 129 130 ACCEPTED MANUSCRIPT 131 MATERIALS AND METHODS 132 Experimental animals and tissue preparation Male inhα/Tag and negative control littermate (C57Bl/6N) mice were used as the animal 134 model Mice were kept in a specific pathogen-free surrounding with controlled light (12 h light, 12 h 135 darkness) and temperature (21±1 ⁰C), fed with mouse chow SDS RM-3 (Witham, Essex, UK) and tap 136 water ad libitum The Ethics Committees for animal experimentation of the Turku University and the 137 State Provincial Office of Southern Finland approved the animal experiments RI PT 133 Prepubertal gonadectomy in inhα/Tag and WT mice was performed at 21-24 days of age 139 Surgery was performed under isoflurane (2‐4%) anesthesia (Isoflo, Orion Pharma, Turku, Finland) and 140 Temgesic (buprenorphine, 0.1 mg/kg/8 h) (Schering-Plough, Brussels, Belgium) was administered as 141 mid- and post-operative analgesia Seven-month-old GDX WT and inhα/Tag, and intact WT mice 142 were sacrificed by exsanguination under isoflurane anesthesia Tissue weights were recorded, adrenals 143 snap-frozen in liquid nitrogen, and/or fixed with 4% paraformaldehyde (PFA) From each mouse 144 about 900µl - 1ml of blood was collected into a tube consisting 100µl of 0,5M sterile EDTA solution, 145 plasma was fractioned by centrifugation at 3000 RPM for 10 minutes in 4°C and stored in -80 °C for 146 further analysis 147 Hormone measurements EP 148 TE D M AN U SC 138 LH levels in plasma were measured by immunofluorometric assay (DELFIA; PerkinElmer) as 150 described previously (Haavisto, et al 1993) Progesterone and testosterone blood plasma 151 concentrations were analyzed by Elecsys® Progesterone II and Testosterone II assays (Roche 152 Diagnostics, Basel, Switzerland), using Cobas e411 immunoanalyzer (Roche Diagnostics, Basel, 153 Switzerland) Detection limits for progesterone and testosterone were 0.10 nmol/l and 0.09 nmol/l 154 respectively 155 Microarray AC C 149 ACCEPTED MANUSCRIPT To identify novel biomarker genes for adrenocortical tumors, we compared adrenals of 7mo 157 GDX inhα/Tag and WT males Total RNA was extracted from frozen tissues using RNeasy Mini Kit 158 (Qiagen, Germantown, MD), re-suspended in 50 µl of nuclease-free water (Promega, Madison, WI), 159 quantified spectrophotometrically (NanoDrop; Thermo Fisher Scientific, Waltham, MA) and then 160 qualified using Bioanalyzer nano kit (Agilent Technologies, Santa Clara, CA) The RNA stock was 161 divided into two batches, in order to use the same template for both microarray and gene expression 162 validation by qPCR RNA was transcribed (n=4/group) using the MessageAmp II aRNA 163 Amplification Kit (Thermo Fisher Scientific) and applied to Agilent whole mouse genome oligo 164 microarrays 4X44K (#GPL7202, Agilent Technologies) accordingly to manufacturer’s protocol 165 (Agilent Technologies) The readout was loaded into the Iimma R/Bioconductor package, normalized 166 between arrays using the quantile normalization After performing a row-wise t-test, fold changes were 167 log2-transformed Genes with fold change higher than 1.5 fold and p-value lower than 0.05 were 168 considered as differently expressed Heatmaps were generated using the gplot package in R The list of 169 significantly differentially expressed genes were uploaded to GOrilla (Eden, et al 2009) and separate, 170 process-based enrichment lists for up- and down-regulated genes were generated 171 Real-time quantitative PCR (qPCR) TE D M AN U SC RI PT 156 Prior to qPCR, 900ng of total RNA was DNaseI treated (Thermo Fisher Scientific) and 173 transcribed (60 in 37°C) using DyNAmo™ cDNA Synthesis Kit (#F470, Thermo Fisher 174 Scientific) qPCR was carried out on a CFX96 Real Time PCR Detection System (BioRad, Vienna, 175 Austria), using the DyNAmo™ Flash SYBR® Green qPCR Kit (#F415, Thermo Fisher Scientific) with 176 15ng of cDNA template in total reaction volume of 20µl, in duplicates Conditions were as follows: 95 177 °C for min, [95 °C for 15 s, 54-62 °C for 15 s, 72°C for 15s] x 40, 72 °C for min, 65-95°C melt 178 curve Sequences of primers are listed in Suppl Table S1 Ct for each gene of interest was normalized 179 by 2-3 reference genes: cyclophilin A (Ppia), β-glucuronidase (Gusb), and hypoxanthine 180 phosphoribosyltransferase (Hprt1), validated for each experiment using Bio- Rad CFX Manager 181 software (BioRad) Gene expression was calculated using qBase MSExcel VBA applet (Hellemans, et 182 al 2007) AC C EP 172 ACCEPTED MANUSCRIPT 183 Immunohistochemistry PFA fixed paraffin embedded adrenal glands from intact and GDX WT, and inhα/Tag 185 (n=4/group) were sectioned ±5 µm and stored in darkness at +4°C Antigens were retrieved in 10µM 186 citrate buffer (pH6), washed in TBS with 0,1% Tween20 (#P1379, Sigma-Aldrich, Saint Louis, US- 187 MO) Primary and secondary antibodies were listed in Suppl Table S2 HRP signal was visualized 188 after 10 incubation with Liquid DAB+ Substrate Chromogen System (Dako, Glostrup, Denmark) 189 Slides were scanned by Pannoramic 250 Slide Scanner (3DHISTECH Ltd., Budapest, Hungary) 190 Densitometric analysis SC RI PT 184 Images from good-quality representative areas (at least 10 from each slide) of GDX inhα/Tag 192 and WT adrenals (n=4/group) were acquired with Pannoramic Viewer (3DHISTECH Ltd.) at 60X 193 magnification and analyzed by Fiji (Schindelin, et al 2012) In brief, Hematoxylin-Diaminobenzidine 194 (H-DAB) color deconvolution was performed and units of intensity occupied by DAB staining (grey) 195 were counted Intensity numbers were transformed into opitcal density (OD) values using formula OD 196 = log(255/Mean intensity) 197 In situ hybridization TE D M AN U 191 We used RNAscope® 2.5 HD Reagent Kit-BROWN (#322300, Advanced Cell Diagnostics) 199 for in situ hybridization (ISH) (Wang, et al 2012) with predesigned probes for Lhcgr, Ppib (control 200 reference probe) and DapB (from Bacillus S., nonsense probe) Hybridization was performed 201 accordingly to manufacturer’s protocol Slides were scanned by Pannoramic Midi FL slide scanner 202 (3DHISTECH Ltd.) 203 Statistical analysis AC C EP 198 204 Numerical data were shown as mean±SEM Graphs and statistical analysis using t-test (for 205 experimental groups) or one-way ANOVA (for groups bigger than 2) followed by the Dunnett's ACCEPTED MANUSCRIPT 206 comparison test were performed using Graph Pad Prism (GraphPad Software, San Diego California, 207 USA) We considered p values