Scholarly article on topic 'Controlled Ovarian Hyperstimulation for In Vitro Fertilization Alters Endometrial Receptivity in Humans: Protocol Effects'

Controlled Ovarian Hyperstimulation for In Vitro Fertilization Alters Endometrial Receptivity in Humans: Protocol Effects Academic research paper on "Biological sciences"

0
0
Share paper
Academic journal
Biology of Reproduction
OECD Field of science
Keywords
{""}

Academic research paper on topic "Controlled Ovarian Hyperstimulation for In Vitro Fertilization Alters Endometrial Receptivity in Humans: Protocol Effects"

BIOLOGY OF REPRODUCTION 82, 679-686 (2010) Published online before print 30 December 2009. DOI 10.1095/biolreprod.109.081299

Controlled Ovarian Hyperstimulation for In Vitro Fertilization Alters Endometrial Receptivity in Humans: Protocol Effects1

Delphine Haouzi,3'4'5'6 Said Assou,3'4'5'6 Clothilde Dechanet,3'4'6 Tal Anahory,4 Herve Dechaud,3'4'5'6 John De Vos,3'5'6 and Samir Hamamah2'3'4'5,6

Institut de Recherche en Biotherapie,3 CHU Montpellier, Hôpital Saint-Eloi, Montpellier, France

ART/PGD Division,4 Departement de Médecine et Biologie de la Reproduction, CHU Montpellier, Hopital Arnaud

de Villeneuve, Montpellier, France

INSERM U847,5 Hopital Saint-Eloi, Montpellier, France

Laboratoire 'Developpement embryonnaire precoce et cellules souches embryonnaires humaines',6 UFR de Medecine, Universite Montpellierl, Montpellier, France

ABSTRACT

The impact of gonadotropin-releasing hormone (GnRH) agonist long compared with GnRH antagonist protocols, under in vitro fertilization conditions on endometrial receptivity, is still debated. Therefore, we compared the effect of both GnRH antagonist and agonist long protocols on the endometrial receptivity by analyzing, to our knowledge for the first time, the global gene expression profile shift during the prereceptive and receptive stages of stimulated cycles under the two GnRH analogue protocols compared with natural cycles in the same patients. For the same normal-responder patients, endometrial biopsies were collected on the day of oocyte retrieval and on the day of embryo transfer after human chorionic gonadotropin administration of a stimulated cycle with either GnRH agonist long or GnRH antagonist protocols and compared with the prereceptive and receptive stages of a natural cycle. Samples were analyzed using DNA microarrays. Gene expression profiles and biological pathways involved during the prereceptive stage to the receptive endometrial transition of stimulated and natural cycles were analyzed and compared for each patient. Both protocols affect endometrial receptivity in comparison with their natural cycle in the same patients. Major differences in endometrial chemokines and growth factors under stimulated cycles in comparison with natural cycles were observed. Such an effect has been associated with gene expression alterations of endometrial receptivity. However, the endometrial receptivity under the GnRH antagonist protocol was more similar to the natural cycle receptivity than that under the GnRH agonist protocol.

cytokines, female reproductive tract, GnRH analogues, gonadotropin-releasing hormone, growth factors, human endometrium receptivity, implantation, microarray, natural cycle

INTRODUCTION

Studies comparing gonadotropin-releasing hormone (GnRH) agonist long with GnRH antagonist protocols remain

1Supported in part by Ferring Pharmaceuticals and Vitrolife. Correspondence: Samir Hamamah, ART/PGD Division, Departement de Medecine et Biologie de la Reproduction, CHU Montpellier, Hôpital Arnaud de Villeneuve, 371 avenue du Doyen Gaston Giraud, 34 295 Montpellier Cedex 5, France. FAX: 33 (4) 67 33 62 90; e-mail: s-hamamah@chu-montpellier.fr

Received: 4 September 2009. First decision: 15 October 2009. Accepted: 10 December 2009.

© 2010 by the Society for the Study of Reproduction, Inc. elSSN: 1529-7268 http://www.biolreprod.org ISSN: 0006-3363

controversial, but data regarding the impact of these protocols on endometrial receptivity also conflict, regardless of approaches to assess endometrial receptivity itself [1-11]. New approaches to the evaluation of endometrial receptivity have been suggested by newly identified molecular biomarkers [1214]. Some studies have provided evidence that both GnRH agonist long and antagonist protocols only lightly affect endometrial receptivity in comparison with natural cycles [2, 15]. By contrast, other studies have suggested that the GnRH antagonist protocol has a strong impact on the expression of genes involved in the human endometrial receptivity [3, 16]. Finally, still other studies have provided evidence that GnRH agonist long protocols produce a delay of endometrial receptivity [3, 5, 17].

The divergence between the results of published studies analyzing the effects of GnRH analogues on endometrial receptivity could be explained by 1) differences in the day of the endometrial biopsies, 2) differences in patient profiles, 3) differences in the protocols used for controlled ovarian hyperstimulation (COS) protocols, 4) inappropriate comparisons between samples, and 5) limited numbers of endometrial samples studied (Table 1). In addition, the study design of these microarray data compared endometrial biopsies between natural and stimulated cycles (after luteinizing hormone [LH] peak on Day 2 or Day 5 (LH+2/5] compared with human chorionic gonadotropin [hCG] 2 days after administraton [hCG+2] or LH+7/9 compared with hCG +7/9) for the same patients without using paired samples [2, 5, 15-18], but doing so is a crucial condition to limit the impact of interpatient variability, as we have shown previously [19].

The aim of the present study was to evaluate the effect of GnRH antagonist versus long protocols on endometrial receptivity by comparing, to our knowledge for the first time, the shift in global gene expression profiles in human endometrial biopsies from normal-responder patients during the early secretory and the midsecretory phases of stimulated cycle with the natural cycle in the same patients (hCG+2 versus hCG+5 compared with LH+2 versus LH+7).

MATERIALS AND METHODS

Patient Characteristics and Endometrial Biopsies

The study population included 30 young, normal-responder patients referred for Intra Cytoplasmic Sperm Injection (ICSI) for male infertility factor between January 2006 and June 2008 and recruited after written informed consent. These patients were part of our previous study comparing the stimulated cycle effects on endometrial receptivity compared to natural cycles without taking into account COS protocols [19]. This project has received

CD Q_ CD Q_

CD T3 —Ï

—Ï (Q

TABLE 1. Study design of seven microarray analyses comparing the prereceptive or receptive stages between natural and GnRH antagonist (Atg) or agonist long (Ag) protocols.

No. of Paired Impact on

Study samples Natural cycle (n)a Stimulated cycle (n)a sample analysis endometrial receptivity

Mirkin et al., 2004 [15] 13 LH+8 (5) hCG+9 Atg (5) No Light

LH+8 (5) hCG+9 Ag (3) No Light

Horcajadas et al., 2005 [17] 19 LH+7 (14) hCG+7 Ag (5) No Strong

Simon et al., 2005 [2] 28 LH+7 (14) hCG+7 Atg standard dose (4) No Light

LH+7 (14) hCG+7 Atg high dose (5) No Light

LH+7 (14) hCG+7 Ag (5) No Light

Horcajadas et al., 2008 [5] 20 LH+2 (5) hCG+2 Ag (5) No Delay

LH+7 (5) hCG+7 Ag (5) No Delay

Liu et al., 2008 [18] 13 LH+7 (5) hCG+7 Ag high serum E2 levels (4) No Strong

LH+7 (5) hCG+7 Ag low serum E2 levels (4) No Strong

Macklon et al., 2008 [16] 8 LH+5 (4) hCG+2 Atg (4) 2 No Strong

Current study 84 LH+2 (7) vs LH+7 (7) hCG+2 Ag (7) vs hCG+5 Ag (7) Yes Strong

LH+2 (14) vs LH+7 (14) hCG+2 Atg (14) vs hCG+5 Atg (14) Yes Strong

a n = number of samples.

institutional review board approval. All patients had normal serum follicle-stimulating hormone, luteinizing hormone (LH), and estradiol on Day 5 and were normal responders during a previous first ICSI attempt under GnRH agonist long protocols (n = 16) (Supplemental Table S1, all Supplemental Data are available online at www.biolreprod.org). However, some whose ICSI failed were, in a second attempt, stimulated under GnRH antagonist protocols (n = 14). A delay of more than 3 mo without treatments elapsed between the two cycles. During the same natural cycle for each patient and before any COS treatment, two endometrial biopsies were done on Day 2 (LH+2) and Day 7 (LH+7) after the LH peak. Because all the patients reported regular cycles, the LH surge was estimated according to the first day of menstruation. Histological analysis was not performed to verify that the LH timing was accurate. Therefore, the possibility for a 1-day delay from the first day of the menstruation cannot be excluded. During COS protocols, as previously described [19], two endometrial biopsies were done, first on the day of egg collection (hCG+2) and another on the day of embryo transfer (hCG+5) after hCG administration. Of the 16 patients treated with GnRH agonist long protocols, nine were not evaluated, because only one endometrial biopsy was performed on hCG+2. Each biopsy sample was frozen at —80°C in RLT RNA extraction buffer (RNeasy Mini kit; Qiagen).

Complementary RNA Preparation and Microarray Hybridization

Total RNA (100 ng) was used to prepare twice-amplified, labeled cRNA for hybridization to HG-U133 plus 2.0 arrays (Affymetrix) as described previously [13, 19]. Each endometrial sample (n = 84) was put individually on a microarray chip.

Data Processing

Scanned GeneChip images were processed using Affymetrix GCOS 1.4 software to obtain an intensity value signal and a detection call (present, marginal, or absent) for each probe set using the default analysis settings and global scaling as first normalization method. Probe intensities were derived using the MAS5.0 algorithm. This algorithm also determines whether a gene is expressed with a defined confidence level or not (detection call). This call can either be ''present'' (when the perfect match probes are significantly more hybridized than the mismatch probes; false discovery rate [FDR] < 0.04), ''marginal'' (0.04 < FDR < 0.06), or ''absent'' (FDR > 0.06). The microarray data were obtained in our laboratory in agreement with the Minimal Information About Microarray Experiment (MIAME; http://www.mged.org/ Workgroups/MIAME/miame.html) recommendations.

Microarray Data Analysis

The Significant Analysis of Microarrays (SAM; Stanford University) [20] was used to identify genes for which the expression varied significantly between the hCG+2 and hCG+5 samples (paired-sample analysis) from patients treated with GnRH agonist long or antagonist protocols and between their respective samples under natural cycles (LH+2 vs. LH+7, paired-sample analysis) as control for each group of protocol. SAM provides mean or median fold-change (FC) values and an FDR confidence percentage based on data permutation. To perform the comparison of gene expression profiles between

endometrial sample groups, a probe set selection using the absent/present detection call (present in at least four and seven samples in the GnRH agonist long and antagonist sample groups, respectively) and a coefficient of variation (>40%) between samples was first performed before the SAM. Selected gene lists (FC > 2 and FDR < 0.05) were submitted to Ingenuity (http://www. ingenuity.com) and FatiGO+ (http://babelomics2.bioinfo.cipf.es) software to identify the biological mechanisms altered by these gene expression variations [21].

We also performed an unsupervised hierarchical clustering of samples with our predictor list of endometrial receptivity as previously described [13]. Hierarchical clustering analysis based on the expression levels of varying probes were performed with the CLUSTER and TREEVIEW software packages [22].

Quantitative RT-PCR Analyses

The RNA (0.35 ig) from endometrial samples (n = 20) was used to generate first-strand cDNA. These cDNAs (2 il of a 1:3 dilution) were used for quantitative PCR reaction according to the manufacturer's recommendation (Applied Biosytems). The 10-il reaction mixture consisted of cDNA (2 il), 2.5 iM primer, and 5 il of LightCycler 480 SYBR Green I Master 2X (Roche). The amplification was measured during 40 cycles with an annealing temperature at 65°C. Quantitative RT-PCR (qRT-PCR) was performed using the LightCycler 480 detection system (Roche) and normalized to PGK1 (phosphoglycerate kinase 1) for each sample using the following formula: E , . . AC7EprlI,,ACt (E = 10—1/slope), where ACt = Ct control - Ct unknown

tested primer PGK1 -

and E corresponds to the effectiveness of the PCR reaction. This effectiveness is obtained by a standard curve corresponding to the primers used. For each hCG+5 sample (n = 5 under GnRH agonist long and n = 5 under GnRH antagonist protocols), the corresponding natural sample (LH+7, n = 10) was used as a control. Each sample was analyzed in duplicate, and multiple water blanks were included with the analysis.

Statistical Analyses

Statistical analyses for qRT-PCR values were performed with SPSS 18.0 software (SPSS, Inc.). A repartition difference between sample groups was considered to be significant when the Kruskal-Wallis nonparametric test gave P < 0.05.

RESULTS

Gene Expression Profiles of Endometrial Receptivity Resulting from GnRH Antagonist Versus Agonist Long Protocols

The SAM analyses were performed between the hCG +2 and hCG+5 sample groups for each protocol (hCG+2 vs. hCG+5; paired-sample analysis) and between the LH+2 and LH+7 control sample groups (LH+2 vs. LH+7; paired-sample analysis). The number of genes significantly up-regulated during the receptive stage was similar between the GnRH

CD Q_ CD Q_

CD T3 —s

—s (Q

FIG. 1. Percentage of genes in common between stimulated and natural cycles during the receptive endometrium. A) The number of genes significantly modulated between the prereceptive (hCG+2) and receptive (hCG+5) samples under GnRH agonist long and antagonist protocols. B) Percentage of genes up- and down-regulated during endometrial receptivity from (left) GnRH agonist, and (right) GnRH antagonist versus their respective natural cycles. C and D) Major down-regulated genes related to the G1/S checkpoint regulation (C) and the G2/M DNA damage checkpoint (D) in the receptive endometrium under GnRH agonist long (dark green) and antagonist (light green) treatments.

agonist long (731 genes) and antagonist (634 genes) protocols. However, there were twice as many down-regulated genes under GnRH agonist long as under antagonist protocols (451 vs. 210 genes, respectively) (Fig. 1A).

The gene lists identified with SAM analyses between the prereceptive and receptive samples from the stimulated cycle under GnRH agonist long or antagonist protocols were then intersected with their respective prereceptive and receptive samples from the natural cycle used as controls to determine their overlap. The number of down-regulated genes in common between stimulated and natural cycles was similar under GnRH agonist long and antagonist protocols (10% in common with their natural cycle under the GnRH agonist protocol vs. 7% under the GnRH antagonist protocol). Of the genes up-regulated during the receptive stage of the natural cycle, 5% matched those up-regulated under GnRH agonist long protocol, whereas 36% were common to the profile resulting from the GnRH antagonist protocol (Fig. 1B).

Differential Gene Expression Profiles Between GnRH Antagonist and Agonist Long Protocols

To identify genes specifically modulated during the prereceptive to the receptive endometrial transition either

under GnRH agonist long or antagonist protocols, we cross-listed the gene lists resulting from each protocol in comparison with their natural cycle. Many genes were commonly regulated under each protocol as compared to the natural cycle (114 up-regulated and 77 down-regulated), but each category displayed a specific gene expression profile. Among down-regulated genes specific to the GnRH agonist signature during the endometrial transition, numerous genes were articulated around cell-cycle function and more especially in checkpoint regulation (Fig. 1C). Among genes specific to this function, the most representative genes were CCNB1 (x-2.2, FDR = 0.005), CCNB2 (x-2.8, FDR = 0.006), CCNE2 (x-3.4, FDR < 0.0001), CCND2 (x-2.3, FDR = 0.005), CDK2 (x-2.1, FDR = 0.005), BRCA1 (x-2.6, FDR = 0.004), CDC2 (x-3.5, FDR = 0.003), CHEK1 (x-2.2, FDR = 0.01), FANCD2 (x-2.6, FDR < 0.0001), RAD51 (x-2.6, FDR = 0.007), and SUV39H1 (x-2, FDR < 0.0001). In contrast, only two down-regulated genes resulting specifically from the GnRH antagonist protocol were associated with cell-cycle function: HDAC9 (x-5.7, FDR < 0.0001) and CDC25C (x-3.6, FDR < 0.0001).

We then performed a FatiGO+ analysis on genes up-regulated during the prereceptive to the receptive endometrial transition and exclusive to either protocol. The majority of the

FIG. 2. Major differences of chemokines and growth factors involved during endometrial receptivity between natural, GnRH antagonist, and GnRH agonist long protocols. Chemokines (A) and growth factors (B) up-regulated during the endometrial receptivity are shown under natural, GnRH agonist long, and GnRH antagonist protocols. Red indicates not present in natural cycles.

up-regulated genes localized as intracellular under the GnRH agonist long protocol and were localized to the extracellular space under the GnRH antagonist protocol (adjusted P < 0.05). Biological processes associated with this list of genes included ''response to external stimulus'' and "defense response'' (adjusted P < 0.05). By contrast, no significant biological processes were associated with the genes up-regulated during the GnRH agonist long protocols. Therefore, for the up-regulated genes specific to each protocol, we focused on the gene families with an extracellular localization that regulate the endometrial microenvironment, such as the chemokines and growth factors.

Major Differences of Chemokines and Growth Factors Between GnRH Agonist Long and Antagonist Protocols

Chemokines such as CCL4, CCL8, CCL14, CCL18, CCL21, and CXCL12 were the most important members of the chemokine family up-regulated during the prereceptive to the receptive endometrial transition and exclusive to the natural cycle, whereas CXCL1, CXCL16, and CCL2 were those resulting only from the GnRH antagonist protocols (vs. CXCL5, CXCL6, and CXCL8 from the GnRH agonist long protocols). Three chemokines were common to the natural and the stimulated cycles under antagonist conditions: CXCL13, CXCL14, and CCL8. Conversely, only one gene, CXCL13, was found in common for the natural cycle and the GnRH agonist long protocol (Fig. 2A). Both FC and FDR were shown in the Supplemental Table S2A. Some of these genes were validated by qRT-PCR (Fig. 3).

Regarding growth factors, FGF7, FGF18, HGF, PDGFA, PDGFRA, TGFA, VEGFA, VEGFB, IGFBP1, IGFBP3, IGFBP7, and IGF2BP2 were the up-regulated genes specific to the natural cycle. IGFBP5 was the sole up-regulated gene specifically expressed under the GnRH antagonist protocol, whereas EGF, GDF15, IGF1R, MET, and IGF2BP3 were up-regulated only under the GnRH agonist long protocol. Two other genes, HBEGF and MEGF10, were found in the stimulated cycles either under the GnRH antagonist or agonist long protocols, but not in the natural cycle. In addition, HGF, PROK1, and TGFB2 were the three genes found exclusively in common between the natural cycle and the GnRH antagonist protocol. TYMP and FGFR2 were the only two genes in common between stimulated and natural cycles (Figs. 2B and 3) (FC and FDR shown in the Supplemental Table S2B).

Top-ranked functional networks around these chemokines and growth factors were also identified for each class of GnRH analogues (Supplemental Fig. S1).

Biomarkers of Endometrial Receptivity Under Stimulated Cycles

Using our previously described predictor list of the endometrial receptivity [13] for unsupervised clustering of the natural and stimulated cycle endometrium profiles from the same patients during the endometrial shift revealed that both GnRH agonist long and antagonist samples were misclassified (Fig. 4). However, the signature of endometrial receptivity was more altered under GnRH agonist long than under GnRH antagonist protocols. Impairment of gene expression of endometrial receptivity was also confirmed by the SAM analyses between the prereceptive and the receptive samples showing that only four and three of our five biomarkers were significantly up-regulated under GnRH agonist long and antagonist protocols, respectively. C2CD4B (NLF2) was increased by a factor of 22.5 in the natural cycle versus 11.5 and 5.4 under GnRH agonist long and antagonist protocols, respectively; a factor of 10.2 versus 9.6 under GnRH antagonist cycles for PROK1; a factor of 37 versus 12.9 under GnRH antagonist cycles for MFAP5; a factor of 12.6 versus 3.8 under GnRH agonist long protocols for ANGPTL1; and a factor of 20.4 versus 8.5 and 12.4 under GnRH agonist long and antagonist protocols, respectively, for LAMB3 (Supplemental Table S2C). However, we found similar variations with a lesser amplitude under stimulated cycles in comparison with natural cycles, suggesting incomplete receptivity during COS protocols. Data obtained from the microarray analyses were previously validated by qRT-PCR [13, 19].

DISCUSSION

The complex processes of embryo implantation require a multitude of molecules acting locally on the endometrium. In the current study, we report different expression patterns of endometrial chemokines and growth factors between GnRH agonist long and antagonist protocols in comparison with natural cycles, affecting the local microenvironment of endometrial receptivity. Such differences were more important under GnRH agonist long than under GnRH antagonist protocols. Our results support previous findings [2] suggesting that GnRH antagonist protocols mimic the natural endometrial receptivity more closely than GnRH agonist long protocols.

Indeed, in the present study, the gene expression profile comparison during the prereceptive to the receptive endome-trial transition revealed that the number of genes in common between stimulated and natural cycles was more important under GnRH antagonist than agonist long protocols (43% vs. 15%). Under the GnRH antagonist protocol, 38% of chemo-kines and 25% of growth factors identified were also observed

1 o 0) Q_ CD Q_

CD T3 —s

—s (Q

FIG. 3. Gene expression validation. Several genes up-regulated during the prereceptive stage to the receptive endometrial transition under either GnRH agonist long (n = 5) or antagonist (n = 5) treatments were validated by qRT-PCR. Bars represent the mean ± SEM. Using the Kruskal-Wallis nonparametric test, differences of mean ± SEM between sample groups were significant at P < 0.05 (*). NS, not significant.

in natural cycles versus 12.5% and 12.5%, respectively, under the GnRH agonist long protocol. Our predictor list of genes related to endometrial receptivity during natural cycles [13] supported such findings.

We reported that among the genes specifically up-regulated under either of the stimulating protocols as compared with their natural cycles, few were in common, suggesting that protocols affect the endometrial receptivity differently. Accordingly, CXCL1, CXCL16, CCL2, and IGFBP5 were exclusively up-regulated during the endometrial transition treated with GnRH antagonists, whereas CXCL5, CXCL6, CXCL8, EGF, GDF15, IGF1R, IGF2BP3, and MET were those exclusively up-regulated under GnRH agonist long protocols. The majority of these findings are reported for the first time, except for the expression of IGFBP5, which was previously reported by Liu

et al. [18] to be up-regulated under GnRH agonist protocols compared to natural cycles. In addition, the MET gene was recently reported by Horcajadas et al. [5] to be up-regulated during the implantation window and delayed in GnRH agonist long protocols on hCG +7 as compared to natural cycles of fertile patients included for the egg donation program. Such disparities between studies could arise from differences in experimental design as well as in microarray data analyses [19]. For this reason, we adopted a new study design strategy for microarray analyses in which the gene expression profile shift from the prereceptive to the receptive stages was compared between stimulated and natural cycles in the same patients.

On the other hand, among these specific gene profile signatures, we also reported a decrease in the expression of genes involved in cell-cycle function during the endometrial

Q_ CD Q_

CD T3 —Ï

—ï (Q

transition. Although we previously reported that treatment with GnRH analogues during COS has a negative impact on cell-cycle function [19], the present study reveals that this phenotype is exclusively associated to the GnRH agonist long protocols. This finding was consistent with several reports showing that GnRH agonist protocols exert an inhibitory effect on the human endometrial cell cycle both in vivo and in vitro [23, 24].

Many in vitro fertilization (IVF) teams currently use GnRH antagonists as a second-choice treatment in normal-responder patients who have failed a previous IVF cycle under GnRH long protocols, but in older patients or in those with polycystic ovary syndrome, the pregnancy rate and live birth rates (per embryo transfer) from patients with equal demographic and clinical characteristics are equivalent between the two GnRH analogues treatments [25, 26]. Consequently, and according to our findings, the COS protocols assume a crucial importance. However, whereas GnRH antagonists cause immediate and rapid gonadotropin suppression by competitive occupancy of the GnRH receptor, GnRH agonists exert their suppression action via pituitary desensitization resulting in a residual gonadotropin activity in some patients.

Little is known regarding the potential effects of LH during the follicular cycle of the menstrual cycle on the endometrial functional activation. On the other hand, preovulatory endometrium is exposed to hCG in most currently used COS protocols, but it is not the case under physiologic conditions. Basically, hCG is used instead of LH to perform final follicle growth and oocyte maturation as well as ovulation control moment. Both hCG and LH bind to a common specific receptor (LH/hCG receptor [hCGR]) belonging to the G protein-coupled receptors family [27]. Recent data suggested that LH or hCG may act directly on the uterus [28, 29]. Tesarik et al. [30] reported that endometrial maturation was disturbed in women treated with GnRH agonists, resulting in low endogenous LH concentrations, but can be rescued by midcycle stimulation of LH/hCGR with exogenous hCG, suggesting that LH/hCGR is needed to support endometrial uterine receptivity. Recently, Kolibianakis et al. [31] reported that the prolongation of the follicular phase by delaying hCG administration results in a higher endometrial advancement incidence, based on histological assessments by Noyes' criteria, on the day of oocyte retrieval in GnRH antagonist cycles [32]. The mechanism by which LH/hCG can directly affect uterine receptivity is still unclear. Certain molecules, such as cytokines (LIF, CSF1, and IL1) and integrins, known to be involved in the endometrial receptivity, may be modulated by LH or hCG [33, 34]. Using an intrauterine microdialysis device to administer low doses of hCG (500 IU/ ml) into the uterine cavity of women during the luteal phase, Licht et al. [29] reported that the intrauterine concentration of IGFBP1 was significantly inhibited during the late luteal phase by hCG administration, but not during the implantation window. However, under COS protocols, hCG administration (5000 or 10 000 IU) was performed 36 h before oocyte retrieval, and to our knowledge, no study to date has evaluated its impact on the gene expression profile during endometrial receptivity.

In addition, the presence of endometrial LH/hCGR is still debated. Several authors have reported the expression of LH/ hCGR in both epithelial and stromal cells of the human endometrium by real-time PCR, nested RT-PCR, and Northern and Western blot analysis [35-39], whereas others were unable to amplify the full-length LH/hCGR mRNA [40]. These disparities seem to be associated with the low level of LH/ hCGR mRNA, making its detection difficult by RT-PCR

FIG. 4. Unsupervised hierarchical clustering of the prereceptive and receptive samples with the predictor list showing common signatures under natural and stimulated cycles. The comparisons of gene expression between natural and stimulated cycles revealed endometrial profiles associated with an altered receptivity during the COS protocol, either under GnRH agonist long or antagonist protocols. PR, prereceptive;R, receptive;pink, up-regulated genes;blue, down-regulated genes.

technology. In addition, a truncated and nonfunctional form of the receptor could be amplified and may be a mechanism regulating LH/hCGR down-regulation [39, 40]. We also validated the low LH/hCGR mRNA expression using qRT-PCR in the prereceptive and receptive endometrial samples under both natural and stimulated cycles, because it was found to be undetectable on microarray chips technologies (data not shown). Although the profile of endometrial LH/hCGR expression correlates with the theoretical timing of the implantation window, its predictive value as a biomarker of endometrial receptivity under IVF conditions is under investigation [41].

Other studies have shown that impairment of endometrial receptivity was caused by high serum estradiol and/or progesterone levels under COS protocols [18, 42]. Although estradiol and progesterone have long been believed to be essential for developing an appropriate endometrial environment for blastocyst implantation and are currently used to maintain the luteal phase during IVF, hormones and growth factors secreted by endometrial cell types are able to modulate their effects. In the LH/hCGR knockout model animals, the normalization of serum estradiol and progesterone levels has been shown to be sufficient to restore a normal endometrial receptivity phenotype, including the ability of the uterus to implant donor blastocysts [43]. These findings underline the

complexity of signaling pathways controlling endometrial receptivity under stimulated cycles.

In conclusion, the present study provides new information relating to the COS protocol effects on endometrial receptivity under IVF cycle. The transcriptomic pattern of endometrial cells in stimulated cycles reveals differential endometrial receptivity both under GnRH agonist long and antagonist protocols in comparison with natural cycles. However, the results obtained in the present study will need confirmation in additional larger studies.

ACKNOWLEDGMENTS

We thank the direction of the University-Hospital of Montpellier for support and the ART team for their assistance during the present study. We also thank Dr. Sophie Perrier d'Hauterive for providing endometrial LH/ hCGR qRT-PCR validations.

REFERENCES

1. Kolibianakis EM, Bourgain C, Albano C, Osmanagaoglu E, Smitz J, Van Steirteghem A, Devroey P. Effect of ovarian stimulation with recombinant follicle-stimulating hormone, gonadotropin-releasing hormone antagonists, and human chorionic gonadotropin on endometrial maturation on the day of oocyte pick-up. Fertil Steril 2002; 78:1025-1029.

2. Simon C, Oberye J, Bellver J, Vidal C, Bosch E, Horcajadas JA, Murphy C, Adams S, Riesewijk A, Mannaerts B, Pellicer A. Similar endometrial development in oocyte donors treated with either high- or standard-dose GnRH antagonist compared to treatment with a GnRH agonist or in natural cycles. Hum Reprod 2005; 20:3318-3327.

3. Ruan HC, Zhu XM, Luo Q, Liu AX, Qian YL, Zhou CY, Jin F, Huang HF, Sheng JZ. Ovarian stimulation with GnRH agonist, but not GnRH antagonist, partially restores the expression of endometrial integrin beta3 and leukemia-inhibitory factor and improves uterine receptivity in mice. Hum Reprod 2006; 21:2521-2529.

4. Rackow BW, Kliman HJ, Taylor HS. GnRH antagonists may affect endometrial receptivity. Fertil Steril 2008; 89:1234-1239.

5. Horcajadas JA, Mínguez P, Dopazo J, Esteban FJ, Domínguez F, Giudice LC, Pellicer A, Simon C. Controlled ovarian stimulation induces a functional genomic delay of the endometrium with potential clinical implications. J Clin Endocrinol Metab 2008; 93:4500-4510.

6. Prapas N, Tavaniotou A, Panagiotidis Y, Prapa S, Kasapi E, Goudakou M, Papatheodorou A, Prapas Y. GnRH antagonists and endometrial receptivity in oocyte recipients: a prospective randomized trial. Reprod Biomed Online 2009; 18:276-281.

7. Van Vaerenbergh I, Van Lommel L, Ghislain V, In't Veld P, Schuit F, Fatemi HM, Devroey P, Bourgain C. In GnRH antagonist/rec-FSH stimulated cycles, advanced endometrial maturation on the day of oocyte retrieval correlates with altered gene expression. Hum Reprod 2009; 24: 1085-1091.

8. Nikas G, Develioglu OH, Toner JP, Jones HW Jr. Endometrial pinopodes indicate a shift in the window of receptivity in IVF cycles. Hum Reprod 1999; 14:787-792.

9. Creus M, Ordi J, Faíbregues F, Casamitjana R, Carmona F, Cardesa A, Vanrell JA, Balasch J. The effect of different hormone therapies on integrin expression and pinopode formation in the human endometrium: a controlled study. Hum Reprod 2003; 18:683-693.

10. Oborna I, Novotny R, Brezinova J, Petrova P, Lichnovsky V, Fingerova H. Changes in the development of uterine pinopodes in steroid hormone supplemented cycles. Physiol Res 2004; 53:423-429.

11. Quinn C, Ryan E, Claessens EA, Greenblatt E, Hawrylyshyn P, Cruickshank B, Hannam T, Dunk C, Casper RF. The presence of pinopodes in the human endometrium does not delineate the implantation window. Fertil Steril 2007; 87:1015-1021.

12. Jones RL, Hannan NJ, Kaitu'u TJ, Zhang J, Salamonsen LA. Identification of chemokines important for leukocyte recruitment to the human endometrium at the times of embryo implantation and menstruation. J Clin Endocrinol Metab 2004; 89:6155-6167.

13. Haouzi D, Mahmoud K, Fourar M, Bendhaou K, Dechaud H, De Vos J, Reme T, Dewailly D, Hamamah S. Identification of new biomarkers of human endometrial receptivity in the natural cycle. Hum Reprod 2009; 24: 198-205.

14. Boomsma CM, Kavelaars A, Eijkemans MJ, Amarouchi K, Teklenburg G, Gutknecht D, Fauser BJ, Heijnen CJ, Macklon NS. Cytokine profiling in

endometrial secretions: a noninvasive window on endometrial receptivity. Reprod Biomed Online 2009; 18:85-94.

15. Mirkin S, Nikas G, Hsiu JG, Diaz J, Oehninger S. Gene expression profiles and structural/functional features of the peri-implantation endometrium in natural and gonadotropin-stimulated cycles. J Clin Endocrinol Metab 2004; 89:5742-5752.

16. Macklon NS, van der Gaast MH, Hamilton A, Fauser BC, Giudice LC. The impact of ovarian stimulation with recombinant FSH in combination with GnRH antagonist on the endometrial transcriptome in the window of implantation. Reprod Sci 2008; 15:357-365.

17. Horcajadas JA, Riesewijk A, Polman J, van Os R, Pellicer A, Mosselman S, Simeon C. Effect of controlled ovarian hyperstimulation in IVF on endometrial gene expression profiles. Mol Hum Reprod 2005; 11:195205.

18. Liu Y, Lee KF, Ng EH, Yeung WS, Ho PC. Gene expression profiling of human peri-implantation endometria between natural and stimulated cycles. Fertil Steril 2008; 90:2152-2164.

19. Haouzi D, Assou A, Mahmoud K, Tondeur S, Reme T, Hedon B, De Vos J, Hamamah S. Gene expression profile of the human endometrial receptivity: comparison between natural and stimulated cycles for the same patients. Hum Reprod 2009; 24:1436-1445.

20. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 2001; 98:5116-5121.

21. Al-Shahrour F, Minguez P, Tarraga J, Montaner D, Alloza E, Vaquerizas JMM, Conde L, Blaschke C, Vera J, Dopazo J. BABELOMICS: a systems biology perspective in the functional annotation of genome-scale experiments. Nucleic Acids Res (Web Server issue) 2006; 34:W472-W476.

22. de Hoon MJ, Imoto S, Nolan J, Miyano S. Open source clustering software. Bioinformatics 2004; 20:1453-1454.

23. Lundkvist O, Bergquist C. Morphological studies of human endometrium during continuous LH-RH agonist treatment. Int J Fertil 1986; 30:65-70.

24. Meresman GF, Bilotas M, Buquet RA, Baranao RI, Sueldo C, Tesone M. Gonadotropin-releasing hormone agonist induces apoptosis and reduces cell proliferation in eutopic endometrial cultures from women with endometriosis. Fertil Steril 2003; 80(suppl 2):702-707.

25. Engel JB, Griesinger G, Schultze-Mosgau A, Felberbaum R, Diedrich K. GnRH agonists and antagonists in assisted reproduction: pregnancy rate. Reprod Biomed Online 2006; 13:84-87.

26. Devroey P, Aboulghar M, Garcia-Velasco J, Griesinger G, Humaidan P, Kolibianakis E, Ledger W, Tomas C, Fauser BC. Improving the patient's experience of IVF/ICSI: a proposal for an ovarian stimulation protocol with GnRH antagonist cotreatment. Hum Reprod 2009; 24:764-774.

27. Ascoli M, Fanelli F, Segaloff DL. The lutropin/choriogonadotropin receptor, a 2002 perspective. Endocr Rev 2002; 23:141-174.

28. Filicori M, Fazleabas AT, Huhtaniemi I, Licht P, Rao ChV, Tesarik J, Zygmunt M. Novel concepts of human chorionic gonadotropin: reproductive system interactions and potential in the management of infertility. Fertil Steril 2005; 84:275-284.

29. Licht P, Russu V, Lehmeyer S, Moll J, Siebzehnrubl E, Wildt L. Intrauterine microdialysis reveals cycle-dependent regulation of endome-trial insulin-like growth factor binding protein-1 secretion by human chorionic gonadotropin. Fertil Steril 2002; 78:252-258.

30. Tesarik J, Hazout A, Mendoza C. Luteinizing hormone affects uterine receptivity independently of ovarian function. Reprod Biomed Online 2003; 7:59-64.

31. Kolibianakis EM, Bourgain C, Papanikolaou EG, Camus M, Tournaye H, Van Steirteghem AC, Devroey P. Prolongation of follicular phase by delaying hCG administration results in a higher incidence of endometrial advancement on the day of oocyte retrieval in GnRH antagonist cycles. Hum Reprod 2005; 20:2453-2456.

32. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950; 1:3-35.

33. Lindhard A, Bentin-Ley U, Ravn V, Islin H, Hviid T, Rex S, Bangsb0ll S, S0rensen S. Biochemical evaluation of endometrial function at the time of implantation. Fertil Steril 2002; 78:221-233.

34. Herrler A, von Rango U, Beier HM. Embryo-maternal signaling: how the embryo starts talking to its mother to accomplish implantation. Reprod Biomed Online 2003; 6:244-256.

35. Reshef E, Lei ZM, Rao CV, Pridham DD, Chegini N, Luborsky JL. The presence of gonadotropin receptors in nonpregnant human uterus, human placenta, fetal membranes, and decidua. J Clin Endocrinol Metab 1990; 70:421-430.

36. Bernardini L, Moretti-Rojas I, Brush M, Rojas FJ, Balmaceda JP. Status of hCG/LH receptor and G proteins in human endometrium during artificial

CD Q_ CD Q_

CD T3 —s

—s (Q

cycles of hormone replacement therapy. J Soc Gynecol Investig 1995; 2: 630-635.

37. Lin J, Lei ZM, Lojun S, Rao CV, Satyaswaroop PG, Day TG. Increased expression of luteinizing hormone/human chorionic gonadotropin receptor gene in human endometrial carcinomas. J Clin Endocrinol Metab 1994; 79:1483-1491.

38. Han SW, Lei ZM, Rao CV. Homologous down-regulation of luteinizing hormone/chorionic gonadotropin receptors by increasing the degradation of receptor transcripts in human uterine endometrial stromal cells. Biol Reprod 1997; 57:158-164.

39. Licht P, von Wolff M, Berkholz A, Wildt L. Evidence for cycle-dependent expression of full-length human chorionic gonadotropin/luteinizing hormone receptor mRNA in human endometrium and decidua. Fertil Steril 2003; 79(suppl 1):718-723.

40. Stewart EA, Sahakian M, Rhoades A, Van Voorhis BJ, Nowak RA. Messenger ribonucleic acid for the gonadal luteinizing hormone/human chorionic gonadotropin receptor is not present in human endometrium. Fertil Steril 1999; 71:368-372.

41. Perrier d'Hauterive S, Berndt S, Tsampalas M, Charlet-Renard C, Dubois M, Bourgain C, Hazout A, Foidart JM, Geenen V. Dialogue between blastocyst hCG and endometrial LH/hCG receptor: which role in implantation? Gynecol Obstet Invest 2007; 64:156-160.

42. Krikun G, Schatz F, Taylor R, Critchley HO, Rogers PA, Huang J, Lockwood CJ. Endometrial endothelial cell steroid receptor expression and steroid effects on gene expression. J Clin Endocrinol Metab 2005; 90: 1812-1818.

43. Rao CV, Lei ZM. Consequences of targeted inactivation of LH receptors. Mol Cell Endocrinol 2002; 187:57-67.

1 o 0) Q_ CD Q_

CD T3 —s

—s (Q