Technical Note pubs.acs.org/jpr
Protein and Steroid Profiles in Follicular Fluid after Ovarian Hyperstimulation as Potential Biomarkers of IVF Outcome Mark M Kushnir,*,†,‡,§ Tord Naessén,⊥ Kjell Wanggren,⊥ Alan L Rockwood,†,‡ David K Crockett,†,‡ and Jonas Bergquist‡,§ †
ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, Utah, United States Department of Pathology, University of Utah School of Medicine, Salt Lake City, Utah, United States § Analytical Chemistry/Department of Chemistry, Biomedical Center and SciLife Laboratory, Uppsala University, Uppsala, Sweden ⊥ Department of Women’s and Children’s Health, Uppsala University, Uppsala, Sweden ‡
S Supporting Information *
ABSTRACT: Controlled ovarian hyperstimulation is performed to assist with generation of multiple mature oocytes for use in in vitro fertilization (IVF). The goal of our study was to evaluate differences in protein and steroid profiles in ovarian follicular fluid (hFF) samples obtained during oocyte retrieval from women undergoing IVF treatment and to identify physiological pathways associated with the proteins. The hFF samples were depleted of abundant proteins, fractionated by ultrafiltration, digested, and analyzed by nano-LC−QTOF. Concentrations of 15 endogenous steroids were determined in the samples using LC−MS/MS methods. The total number of proteins identified in the samples was 75, of which 4, 7, and 2 were unique to the samples from women with viable pregnancy, miscarriage, and no pregnancy, respectively. Identified proteins were associated with the acute response signaling, coagulation system, intrinsic and extrinsic prothrombin activation, complement system, neuroprotective role of THOP1, FXR/RXR activation, role of tissue factor, and growth hormone pathways. A greater number of proteins associated with biosynthesis was found in hFF samples corresponding to the oocytes resulting in pregnancy. The abundance of seven proteins was found to be associated with steroidogenesis. The obtained data will contribute to better understanding of the pathogenesis and development of noninvasive markers for assessment of oocytes viability. KEYWORDS: follicular fluid, mass spectrometry, proteins, steroids, IVF
■
INTRODUCTION Controlled ovarian hyperstimulation is performed to assist with generation of multiple mature oocytes for use in in vitro fertilization (IVF). The stimulation overrides the natural physiologic mechanisms that normally select a single dominant follicle per menstrual cycle and involves use of exogenous gonadotropins to promote growth of the follicles and stimulate growth of multiple follicles simultaneously.1−3 Ovarian follicular fluid (hFF) is produced during follicular growth and provides the microenvironment for oocyte development. Main origins of hFF are circulating blood, which diffuses through the follicular wall into an antrum of a follicle, and follicular secretions, which are predominantly products of metabolic processes within an oocyte.1,2 These metabolic products could provide information on the status of “health” of the oocyte and potentially could become diagnostic biomarkers. The composition of hFF is very complex and among other biological molecules it contains regulatory proteins that affect follicular growth, regulate permeability of follicular wall, follicular maturation and ovulation.4−6 Although a large number of proteins has been identified and numerous studies have reported protein profiles in hFF, the pathogenesis and © 2012 American Chemical Society
mechanism of follicular development are still not completely understood.4−6 In our earlier publication, we identified 69 proteins in a pooled sample of hFF fluid samples obtained during oocyte retrieval from women undergoing IVF treatment.7 However, understanding of the functions of many of the proteins and their association with the outcome of the IVF treatment is limited and analysis of individual samples should be highly advantageous. Better characterization of the hFF secretome will not only expand the knowledge of the folliculogenesis and advance understanding of the protein’s role during oocyte development, but also potentially contribute to the development of new methods of assessment of oocytes viability for IVF treatment. Due to the complexity of the composition of hFF, sensitive and specific techniques for identification of the proteins in hFF are essential. To overcome some of the limitations of earlier utilized approaches for analysis of proteins in hFF, we utilized depletion of abundant proteins, prefractionation by molecular weight, in-solution digestion, followed by nano-LC−QTOF Received: June 15, 2012 Published: September 18, 2012 5090
dx.doi.org/10.1021/pr300535g | J. Proteome Res. 2012, 11, 5090−5100
Journal of Proteome Research
Technical Note
Table 1. Anthropometric and Reproductive Characteristics of Women mean [range] variable Age (years) Height (cm) Weight (kg) BMI (kg/m2) Current smokers Parity Number of retrieved eggs Number of embryos top quality Number of the IVF attempts Born baby/Miscarriage/No pregnancy Total dose FSH Number of days with FSH hCG at day of fertilization Concentration of estradiol, ng/mL Concentration of testosterone, ng/mL Concentration of 17OH progesterone, ng/mL Endometriosis Male factor Male and female factor Anovulation Unknown a
baby born (group 1) 30 (24−33) 166 (158−172) 70 (52−81) 25.5 (18.6−32.4) 0 out of 3a 1 woman, 2; 3 women −0 9.25 (3−13) 4 (1−6) All women, 1 4/0/0 151 (105−200) 10.7 (9−12) 9.5 (10−12) 117 (110−138) 0.04 (0.02−0.07) 830 (648−994) Cause of infertility na 2 na 2 na
miscarriage (group 2)
no pregnancy (group 3)
32 (26−38) 166 (159−180) 82 (55−112) 28.8 (21.8−34.6)a 3 out of 3a 1 woman, 1; 3 women, 0 11 (11−11) 2.5 (0−7) 2 women, 3; 2 women, 1 2/4/1 144 (125−150) 11.5 (10−13) 11.2 (10−12) 113 (78−146) 0.27 (0.05−0.31) 1073 (616−1755)
36 (31−39) 167 (156−175) 70 (59−85) 25 (23.3−27.8) 1 out of 4 All women, 0 6.25 (2−9) 2.5 (0−6) 3 women, 1; 1 woman, 2 0/0/5 200 (100−300) 12.7 (12−14) 12.7 (12−14) 136 (11−162) 0.16 (0.05−0.36) 1697 (626−2910)
na 2 na na 2
1 1 1 na 1
Information related to one of the participants is missing.
subcutaneous injection of hCG (Pregnyl, Organon, Oss, The Netherlands). Follicle aspiration was performed by transvaginal needle aspiration under ultrasonographic guidance 35− 36 h after hCG administration.
analysis, and protein identification by a database search of the collected MS/MS data. In this study, we aimed to investigate whether potential protein biomarkers exist in follicular fluid, which are associated with the outcome of IVF treatment, and to determine the physiological pathways, associated with the identified proteins, and the IVF treatment outcomes. Lastly, to complement the protein identifications, the hFF samples were quantitatively analyzed for 20 steroids of the pathway of steroids biosynthesis and association of the abundance of the identified proteins with concentrations of steroids was evaluated.
Follicle Fluid Samples
The stimulated hFF samples used in this study were pooled from multiple aspirated follicles at ovum pick-up of each participating woman. Outcomes of the first embryo transfer in each woman were grouped in three types: Group 1, born baby; Group 2, miscarriage; and Group 3, no pregnancy. Each group consisted of four hFF samples from four women. All samples were stored frozen at −70 °C until use and thawed immediately before the experiments. Samples were transported between participating centers on dry ice. The study was approved by the ethic committee of the Uppsala University (Uppsala, Sweden).
■
MATERIALS AND METHODS Stimulated hFF was sampled from 12 women undergoing IVFtreatment at Uppsala University hospital (Uppsala Sweden), with infertility caused by male factor, anovulation, endometriosis or unexplained infertility. Anthropometric and reproductive characteristics of the women are listed in Table 1.
Reagents and Supplies
Trypsin was purchased from Princeton Separations (Adelphia, NJ). Trifluoroethanol, dithiothreitol, formic acid were obtained from Sigma-Aldrich (St. Louis, MO). C18-SB-Zorbax Chip (40 nL trap, 75 μm × 150 mm, 5 μm particles) and MARS depletion column and reagents kit for removal of six abundant proteins (albumin, transferrin, haptoglobin, IgG, IgA, and alpha-1 antitrypsin) were purchased from Agilent Technologies (Santa Clara, CA). Ultrafilters Amicon Ultra-4 with molecular weight cutoff of 50 kDa were purchased from Millipore (Billerica, MA). SPE columns Strata X were purchased from Phenomenex (Torrance, CA). All other reagents used were of highest purity commercially available. Solvents were of HPLC grade, purchased from JT Baker (Phillipsburg, NJ).
Ovarian Stimulation Protocols
Down-regulation for ovarian hyperstimulation was achieved by using a gonadotropin-releasing hormone (GnRH) agonist (Suprecur, Hoechst, Frankfurt, Germany or Synarela, Syntex Nordica AB, Södertälje, Sweden), starting on cycle day 21, or by GnRH antagonist (Orgalutran, Organon, Oss, The Netherlands) started on stimulation day 6. Ovarian stimulation was induced by using either, rFSH, (Gonal-F, Serono Laboratories, Aubonne, Switzerland, or Puregon, Organon, Oss, The Netherlands) or hMG (Menopur, Saint-Prex, Switzerland). The starting dose was dependent on the woman’s age and/or previous response to ovarian stimulation. The ovarian response was monitored by means of serum estradiol assays and vaginal ultrasonographic scans of follicles. FSH or hMG was administered until the leading follicle had a diameter of at least 17 mm. Maturation of the oocytes was triggered by a
Sample Preparation
Aliquots of hFF samples (20 μL) were depleted of abundant proteins using MARS depletion column according manufacturer recommendations. Depleted samples were concentrated 5091
dx.doi.org/10.1021/pr300535g | J. Proteome Res. 2012, 11, 5090−5100
Journal of Proteome Research
Technical Note
and desalted by filtering through 5 KDa ultrafilter; 2 mL of ultrapure water were added to the fractions retained on top of the filters, and solutions were filtered to the final retained volume of 100 μL. To the resulting samples 800 μL of ultrapure water was added, the samples were filtered through ultrafilters with cutoff of 50 kDa and both fractions (above and below 50 kDa) were retained for analysis. The samples were evaporated using evaporative centrifuge to the final volume of 50 μL, proteins were denatured with trifluoroethanol and reduced with dithiothreitol, cysteines were alkylated with iodoacetamide followed by in-solution digestion with trypsin at 37 °C for 18 h. The samples were dried and reconstituted in 20 μL of solvent.
ppm (for over 80% of the peptides error was within 10 ppm); (c) good alignment of predicted vs observed ion series; (d) number of unmatched ions in mass spectra less than 50% of the expected number of ions; (e) forward score greater than the reverse score; (f) well-defined chromatographic peaks at the retention time where the mass spectrum was acquired. The composite results from individual samples were then compared using the Spectrum Mill software, and lists of proteins and peptides, as well as quantitative information were exported for further evaluation. Proteins were quantitated using spectral counting, where the relative abundance of a protein was assumed proportional to the number of MS/MS spectra acquired from a protein.16 Pathway analysis was performed using Ingenuity Pathway Analysis (IPA) v8.7 (Ingenuity, Redwood City, CA).
Instrumental Analysis
Samples were analyzed on the Agilent 6510 Q-TOF equipped with a ChipCube and series 1200 nano-HPLC system (Agilent Technologies, Santa Clara, CA). Peptide separation was performed on a C-18SB-Zorbax Chip. Mobile phase consisted of 5% acetonitrile in water containing 0.1% formic acid, and 5% water in acetonitrile containing 0.1% formic acid. An aliquot of each sample (0.5 μL) was injected onto loading trap of the chip at flow rate of 4 μL/min with the effluent directed to waste. After one minute, the flow was diverted to the analytical column and acquisition was started. Mobile phase for the analytical separation was delivered at a flow rate of 0.4 μL/min, with a gradient of 5% to 60% acetonitrile in 15 min, followed by conditioning and re-equilibration to initial conditions. Data acquisition (2 GHz extended dynamic range) was performed with the MassHunter software rev.B.02 (Agilent Technologies, Santa Clara, CA) with an acquisition rate of 3 scans/s followed by MS/MS scans of the five most intense ions with the exclusion set for 60 s after five consecutive MS/MS scans. Each sample was analyzed three times using different mass ranges for the precursor selection (m/z 400−620 Da, 620−800 Da and 800−1600 Da). The Q-TOF analyzer was tuned to a resolution of 12 000 and calibrated before each experiment for a mass accuracy of 11). The error rate (as approximated using a scrambled database) for the identification of peptides with score above 11 was