ABSTRACT: Goal of the experiment: To examine differential gene expression in granulosa cells of women undergoing ovarian stimulation with either recombinant human follicle stimulating hormone or purified human menopausal gonadotropin for in vitro fertilization. Brief description of the experiment: The study was designed to test the hypothesis that human granulosa cell (GC) gene expression response differs between pure recombinant FSH and human menopausal gonadotropin (hMG) stimulation regimens. Women < 35 years-old undergoing in vitro fertilization for tubal or male factor infertility were prospectively randomized to one of two stimulation protocols, Gonadotropin releasing hormone (GnRH) agonist long protocol plus individualized dosages of (1) recombinant follicle stimulating hormone (rFSH) (Gonal-F®) (n=4) or (2) purified human menopausal gonadotropin (hMG; Menopur®; 75 IU FSH/75 IU LH per vial) (n=3). Follicle development was monitored by ultrasound and serum estradiol levels. When two follicles were ≥ 17mm, human chorionic gonadotropin (hCG) (10,000 IU; Novarel®) was administered. Oocytes were retrieved 35 h post-hCG. Following oocyte recovery, GC were immediately collected in RNALater®. Pooled GC from individual patients were washed, centrifuged over 40% Percoll®, resuspended in RNALater, and snap-frozen in LN. Total RNA was extracted and stored at -70C. Biotinylated cRNA was synthesized from total RNA and each sample was run individually on CodeLink Whole Human Genome Bioarrays (Applied Microarrays). Unnamed genes and genes with <2-fold difference in expression were excluded from further analysis. After exclusions, 1736 genes exhibited differential expression between groups. Over 400 were categorized as signal transduction genes, ~180 as transcriptional regulators, and ~175 as enzymes/metabolic genes. Expression of selected genes was confirmed by RT-PCR. Differentially expressed genes included A kinase anchor protein 11 (AKAP11), bone morphogenic protein receptor II (BMPR2), epidermal growth factor (EGF), insulin-like growth factor binding protein (IGFBP)-4, IGFBP-5, basigin, and hypoxia-inducible factor (HIF)-1 alpha. The results suggest that major differences exist in the mechanism by which pure FSH alone versus FSH/LH regulate gene expression in preovulatory GC that could impact oocyte maturity and developmental competence. Experimental design: Women < 35 years-old undergoing in vitro fertilization for tubal or male factor infertility were prospectively randomized to one of two stimulation protocols, Gonadotropin releasing hormone (GnRH) agonist long protocol plus individualized dosages of (1) recombinant follicle stimulating hormone (rFSH) (Gonal-F®) (n=4) or (2) purified human menopausal gonadotropin (hMG; Menopur®; 75 IU FSH/75 IU LH per vial) (n=3). Follicle development was monitored by ultrasound and serum estradiol levels. When two follicles were ≥ 17mm, human chorionic gonadotropin (hCG) (10,000 IU; Novarel®) was administered. Oocytes were retrieved 35 h post-hCG. Following oocyte recovery, GC were immediately collected in RNALater®. Pooled GC from individual patients were washed, centrifuged over 40% Percoll®, resuspended in RNALater, and snap-frozen in LN. Total RNA was extracted and stored at -70C. Biotinylated cRNA was synthesized from total RNA and each sample was run individually on CodeLink Whole Human Genome Bioarrays (Applied Microarrays Quality control steps: The cRNA that was synthesized from the granulosa cells from individual patients was used for hybridization to a single CodeLink (Applied Microarrays) Whole Human Genome microarray. Only one sample was hybridized with each slide and only one dye (Alexa 647) was used so no dye swaps were necessary. Bacterial control spikes were used as per manufacturer's instructions. Samples used, extract preparation and labelling: The origin of each biological sample: Granulosa cells from women < 35 years-old undergoing in vitro fertilization. Manipulations of biological samples and protocols used: Female patients were treated with gonadotropin releasing hormone (GnRH) agonist long protocol plus individualized dosages of (1) recombinant follicle stimulating hormone (rFSH) (Gonal-F®) (n=4) or (2) purified human menopausal gonadotropin (hMG; Menopur®; 75 IU FSH/75 IU LH per vial) (n=3). Follicle development was monitored by ultrasound and serum estradiol levels. When two follicles were ≥ 17mm, human chorionic gonadotropin (hCG) (10,000 IU; Novarel®) was administered. Oocytes were retrieved 35 h post-hCG. Following oocyte recovery, granulosa cells were immediately collected in RNALater®. Pooled GC from individual patients were washed, centrifuged over 40% Percoll®, resuspended in RNALater, and snap-frozen in LN. Experimental factor: treatment with recombinant FSH versus purified human menopausal gonadotropin Technical protocols: Granulosa cells were collected into RNAlater (Ambion) and stored at -80C until processed. The granulosa cells were pelleted by centrifugation to remove the RNAlater. Each cell pellet was homogenized in TRI reagent (Molecular Research Centers). Bromochloropropane and sodium acetate were added, and the samples were centrifuged to separate the phases. The RNA-containing layer was removed and the RNA purified on an RNeasy extraction column (Qiagen). The sample was treated with an on-column DNase treatment (RNase-free DNase, Qiagen). The purity and quantity were evaluated by an Agilent Bioanalyzer using the RNA 6000 Nanoassay LabChip. Labelled cRNA was prepared using the manufacturer's Instruction Protocol (www1.amershambiosciences.com, CodeLink Gene Expression System: Manual Labelled cRNA Target Preparation). The source of reagents was the CodeLink Expression Assay Reagent Kit, Manual Prep, except where specified otherwise. 0.285 micrograms of total granulosa cell RNA was mixed with bacterial control RNA spikes and primed with T7 oligo(dT) primer for 10 min at 70C. (The bacterial control spikes included araB, entF, fixB, gnd, hisB, and leuB.) The first strand of cDNA was synthesized with first strand buffer, dNTP mix, RNase inhibitor, and reverse transcriptase for 2 h at 42C. The second strand cDNA synthesis reaction was prepared using second strand buffer, dNTP mix, DNA polymerase mix, and RNase H; the reaction was carried out for 2h at 16C. The double-stranded cDNA was purified on QIAquick columns (Qiagen) and the eluent was dried down in a SpeedVac concentrator. The double-stranded cDNA was resuspended in a mixture containing T7 reaction buffer, T7 ATP, T7 GTP, T7 UTP, T7 CTP, biotin-11-UTP, and T7 enzyme mix for the synthesis of cRNA. The cRNA synthesis reaction was terminated after 14h at 37C by purifying the cRNA on RNeasy columns (Qiagen). The concentration of cRNA was determined by spectrophotometry. Hybridization procedures and parameters: 10 micrograms of cRNA was mixed with fragmentation buffer and heated to 94C for 20 min. The fragmented cRNA was mixed with CodeLink hybridization buffer, loaded on the microarray slides, and hybridized for 18 hours at 37C. The slides were washed in 0.75x TNT (Tris-HCl, NaCl, Tween-20) at 46C for 1h then incubated with streptavidin-Alexa 647 fluorescent dye for 30 min at room temperature. The Alexa fluor was prepared in TNB blocking buffer (0.1M Tris-HCl, 0.15M NaCL, 0.5% NEN Blocking Reagent-PerkinElmer) The slides were then washed 4 times for 5 min each in 1x TNT and once in 0.05% Tween 20 for 5 sec. The slides were dried by centrifugation and scanned in an Axon GenePix 4000B scanner.
