Human Reproduction Update Advance Access originally published online on September 7, 2006
Human Reproduction Update 2007 13(1):77-86; doi:10.1093/humupd/dml046
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Wide genomic analysis of human endometrial receptivity: new times, new opportunities
Fundación Instituto Valenciano de Infertilidad (FIVI), Instituto Universitario IVI, Valencia University, Valencia, Spain
1 To whom correspondence should be addressed at: Fundación IVI, C/Guadassuar, 1, E-46015 Valencia, Spain. E-mail: csimon{at}ivi.es
 |
Abstract |
|---|
TOP
Abstract
IntroductionMicroarray technology has broadened the insight into many research fields allowing scientists to analyse the expression of many genes in quick and efficient experiments aimed at translating these findings into clinical applications. In reproductive medicine, researchers have exploited microarrays to increase understanding of the molecular mechanisms involved in endometrial receptivity and how a possible therapeutic translation can be feasible. In the last 4 years, several studies have focused on the genomics of the human endometrium in different physiological and pathological conditions, and these studies have generated a large amount of information about the regulation and dysregulation of the window of implantation (WOI) genes in fertile, subfertile and refractory conditions. However, the key molecules/mechanisms in endometrial receptivity remain to be elucidated. In this comprehensive review, we have analysed the available results obtained in our own and other laboratories, defining the genomic profile of the receptive endometrium in different situations and its possible clinical application.
Key words: endometrium / gene expression profile / genomics / microarray
|
Introduction |
|---|
The endometrium is a highly dynamic tissue empowered with the capacity to undergo dramatic changes in response to steroid hormones, ultimately aiming to create a window of receptivity for blastocyst implantation. During decades, to understand the basic mechanisms implicated in endometrial receptivity, researchers have investigated the molecular events in the endometrium using the preferred molecular approach in animal and/or human models. The development of the microarray technology (Schena et al., 1995
) has made it possible to analyse the expression of thousands of genes at the same time in a specific sample. This technology allows us to address, in a global manner, the genomics of endometrial development.
In mice, the endometrial tissue has been analysed using this technology in natural cycles (Yoshioka et al., 2000; Reese et al., 2001; Tan et al., 2003) or in response to hormones (Cheon et al., 2002; Curtis-Hewitt et al., 2003; Watanabe et al., 2003; Ho Hong etal., 2004). Similar studies have been carried out in rats (Naciff et al., 2002; Wu et al., 2003), rhesus monkeys (Ace and Okulicz, 2004; Tynan et al., 2005) and cows (Ishiwata et al., 2003), which are discussed in a comprehensive review (White and Salamonsen, 2005). With a variety of in vitro and in vivo models, the gene expression profile of the human endometrium has also been investigated during decidualization (Popovici et al., 2000; Brar et al., 2001; Tierney et al., 2003) or in response to progesterone (Okada et al., 2003). In pathological conditions such as in endometrial cancer, this technical approach has been used to get more insight into the molecular pathways involved (Mutter et al., 2001; Risinger et al., 2003; Moreno-Bueno et al., 2003a; Cao et al., 2004; Ferguson et al., 2004, 2005; Saidi et al., 2004).
Nowadays, compelling evidence exists regarding the identification of endometrial gene expression profiles that identify different phases of the menstrual cycle (Ponnampalam et al., 2004; Punyadeera et al., 2005; Talbi et al., 2005). Even the differential gene expression pattern between the epithelial and stromal compartments during the proliferative (PE) phase of the menstrual cycle has been reported using a combination of microarray technology and laser capture microdissection in humans (Yanaihara et al., 2005) and mice (Niklaus and Pollard, 2006).
Finally, the genomics of human endometrial receptivity has also been explored in natural cycles (Carson et al., 2002; Kao et al., 2002; Borthwick et al., 2003; Riesewijk et al., 2003; Mirkin et al., 2005), in controlled ovarian stimulated cycles (Mirkin et al., 2004; Horcajadas et al., 2005; Simón et al., 2005) and in endometriosis (Eyster et al., 2002; Arimoto et al., 2003; Kao et al., 2003; Matsuzaki et al., 2004, 2005). Because the refractory endometrium represents the opposite part of the spectrum, researchers have investigated the gene expression profile in conditions that will render a receptive endometrium unreceptive, such as the presence of an inert intrauterine device (IUD) (Horcajadas et al., 2006) or endometrial explants treated in vitro with RU486 (Catalano et al., 2003). These results have been partially discussed in previous reviews (Giudice, 2003; Giudice, 2004; Horcajadas et al., 2004a,b), but the extent of recent evidence (Table I) suggests that the time is right for a new objective review.
|
|
Why is a reappraisal of uterine receptivity assessment needed? |
|---|
For more than 50 years, histological evaluation of the endometrium has been the gold standard for clinical diagnosis set on the basis of the morphological observations of Noyes and collaborators (Noyes et al., 1950
, 1975). The authors described ‘specific’ morphological appearance of the different compartments of the endometrium throughout the menstrual cycle. They distinguished different phases of the cycle on the basis of the histological features and even the endometrial dating during the luteal phase. These classic papers have been the most quoted in our field, cited thousands of times and followed as ‘the’ diagnostic tool for endometrial dating and endometrial research worldwide. However, its accuracy and the functional relevance of this system based exclusively on histological observations as predictor of receptivity have been questioned in recent randomized studies (Coutifaris et al., 2004; Murray et al., 2004). Murray and colleagues have demonstrated that histological features solely have high interobserver, intraobserver and intercycle variations (Murray et al., 2004). The differences between pathologists depend on the day of the menstrual cycle when the endometrial biopsy is collected. Also, this methodology does not have the capability to discriminate between fertile and infertile couples (Coutifaris et al., 2004). These studies suggest that new technologies should be added for objective identification of endometrial samples and the study of endometrial development in health and disease.
The group of Peter Rogers was the first to propose the genomic characterization of the human endometrium throughout the menstrual cycle using microarray technology (Ponnampalam et al., 2004). They conclude that it is possible to classify the endometrium precisely according to their transcriptional profile regardless of the morphological appearance. More importantly, they established the existence of clusters of genes characteristic of the different phases of the cycle, highlighting the potential of gene expression profiling for the development of molecular tools in the evaluation of the endometrial status. This study has been confirmed and extended by the group of Linda Giudice that has dissected the molecular phenotyping of human endometrium throughout the menstrual cycle phases underlying its biological processes in normo-ovulatory women (Talbi et al., 2005). Their study results underscore the potential of gene expression profiling for developing molecular diagnostics of normal versus abnormal endometrium and identifying molecular targets for therapeutic purposes in endometrial disorders.
|
Microarray technology and experimental designs |
|---|
DNA microarray technology is so far one of the most widely used and potentially revolutionary research tools derived from the human genome project (Celera Genomics, 2001
). This technique has been developed within the last decade and allows the assessment of the complete genomic expression profile in a given biological sample in a single experiment (Schena et al., 1995). This technological capacity has opened new avenues of biological investigation and generated a large number of applications (Stoughton, 2005).
The technique is based on the complementarity of the DNA duplex and the capability of the single-stranded DNA to bind to solid supports such as nylon membranes or glass. Usually, immobilized probes are hybridized with labelled cDNAs. This labelling can be carried out with fluorescence or radioactivity depending on the support chosen. There is a very wide range of microarrays commercially available separated into two categories: cDNA arrays and high-density synthetic oligonucleotide microarrays (Barret and Kawasaki, 2003). cDNA arrays are much cheaper and easier to produce ‘at home’ and allow the use of dual colour conditions on glass. With this double labelling, it is possible to compare two samples in the same array. Control and experimental cDNAs compete in the same array for hybridizing with a specific probe. High-density synthetic oligonucleotide microarrays are manufactured by several companies, such as Affymetrix or Amersham Biosciences and provide greater hybridization specificity. The main disadvantage of this system is the higher cost. Microarrays, in the two different versions, are now a widely recognized tool for research, diagnosis and therapeutics (Hoheisel, 2006).
The diversity of platforms and analytical methods available to researchers has made the comparison of data from multiple platforms challenging. One group compared different commercial and ‘in-house’ platforms. They found a good consistency for highly expressed genes. Concordance of measurements was higher between laboratories on the same platform than across platforms. They also demonstrated that after stringent preprocessing, commercial arrays were more consistent than in-house arrays, and by most measures, one-dye platforms were more consistent than two-dye platforms (Kuo et al., 2006).
The experimental design, sample collection and selection are crucial when microarray methodology is used (Churchill, 2002). It is also important to consider the biological variation between samples and the inherent troubleshooting of the technique (RNA degradation, quality of labelling, hybridization efficiency, etc.). There are some parameters that have to be annotated for genomic modification of the endometrium such as medical and surgical background (e.g. endometriosis) and concurrent medications (e.g. contraceptive steroids, hormonal replacement therapy and non-steroidal anti-inflammatory inhibitors, among others) that may affect endometrial gene expression. Tissue samples that do not comply with all the requirements indicated must be removed to avoid false interpretations and artefacts. Designing a sound and competent microarray experiment that will obtain sufficient statistical power requires input from a scientific expert with experience in this technology. In general, researchers should avoid pooling RNA, except for experiments that allow the use of a high number of samples and then pool the RNA without loosing information (Kendziorski et al., 2003; Peng et al., 2003). This strategy reduces the number of the required microarrays and the financial cost of the experiment (Kendziorski et al., 2003; Peng et al., 2003), but it is not always possible to collect a high number of samples in practically all the studies in humans. Finally, a reliably well-designed experiment should include an independent validation (Rockett and Hellman, 2004) using other techniques such as quantitative PCR or in situ hybridization.
Although many significant results have been derived from microarray studies in the last decade, one of the most important limitations has been the lack of standard protocol for presenting and exchanging such data. As an interesting initiative, some researchers have introduced a new technique, i.e. the Minimum Information About a Microarray Experiment (MIAME), that describes the minimum information required to ensure that microarray data can easily be interpreted and the results derived from its analysis can be independently verified (Brazma et al., 2001; Parkinson et al., 2005). The ultimate goal is to establish a standard for recording and reporting microarray-based gene expression data, which will in turn facilitate the establishment of databases and public repositories and enable the development of data analysis tools (http://www.mged.org/Workgroups/MIAME/miame.html).
It is important to point out that the main limitation of this technology is that this information represents the transcriptional level, i.e. how the genomic machinery transcribes the genes into RNA. Post-transcriptional events such as protein translocation cannot be assessed within this approach (Mata et al., 2005).
|
Actual perspectives in the interpretation of microarray analysis |
|---|
Microarray technology has experienced a rapid evolution as well as the development of analytical systems. A considerable number of statistical platforms that help the researcher to interpret the results in terms of their biological implications rather than a mere comparison of lists of genes have been developed (Al-Shahrour and Dopazo, 2005
; Al-Shahrour et al., 2005).
One of the most common platforms used is the Gene Expression Profile Analysis Suite (GEPAS), which has been running for >4 years. GEPAS has been designed to provide an intuitive and powerful web–based interface that offers diverse analysis options from the early step of preprocessing (normalization of Affymetrix and two-colour microarray experiments and other preprocessing options) to the final step of the functional annotation of the experiment (using Gene Ontology, pathways, PubMed abstracts and the like) and includes different possibilities for clustering, gene selection, class prediction and array comparative genomic hybridization management. GEPAS is extensively used by researchers of many countries, and its records indicate an average usage rate of 400 experiments per day. The web-based pipeline for microarray gene expression data, GEPAS, is available at http://www.gepas.org (Montaner et al., 2006).
Other platform commonly used by researchers and bioinformatics specialists is the microarray software suite TM4. The TM4 suite of tools consists of four major applications, Microarray Data Manager (MADAM), TIGR_Spotfinder, Microarray Data Analysis System (MIDAS) and Multiexperiment Viewer (MeV), as well as a MIAME-compliant MySQL database, all of which are freely available to the scientific research community at TIGR’s software download site (http://www.tm4.org). Although these software tools were developed for spotted two-colour arrays, many of the components can be easily adapted to work with single-colour formats such as filter arrays and GeneChipsTM (Affymetrix).
There is other very powerful free data annotation software such as DAVID, DRAGON, EASE, FANTOM, GoMiner, MatchMiner, Onto-Express or RESOURCERER that provide a highly scalable platform for enterprise-level genomic research. A comparison among the different softwares deserves a review itself.
Although there are still some concerns with the cross-platform coherence of results but it seems clear that intra-platform reproducibility is high (Moreau et al., 2003) and despite the fact that gene-by-gene results are not always the same, the biological themes emerging from the different platforms are increasingly consistent (Bammler et al., 2005).
|
Genomic analysis of the human endometrium throughout the menstrual cycle |
|---|
Two independent studies have demonstrated the genomic identification of different phases of the menstrual cycle based on the profiling of gene clusters that are expressed in a stage-specific manner (Ponnampalam et al., 2004
; Talbi et al., 2005). They have also demonstrated a strong correlation between the genomic and the histopathological classification. From a total of 425 genes, by hierarchical clustering they found seven clusters of genes specifically up-regulated in a given phase of the cycle (Ponnampalam et al., 2004).
Talbi and colleagues have examined the biochemical molecular signatures of the human endometrium to prove their possible application as a molecular phenotyping tool (Talbi et al., 2005). Principal component analysis (PCA) of their data revealed that samples were self-clustered into four groups that are consistent with histological phenotypes of PE, early secretory (ESE), mid-secretory (MSE) and late-secretory (LSE) phases. They also performed independent hierarchical clustering analysis that revealed similar results, with two major dendrogram branches corresponding to PE/ESE and MSE/LSE that are sub-branched into the four respective phases with heterogeneity among samples within each sub-branch (Talbi et al., 2005). They found four major patterns of gene expression (high in PE, high in ESE, high in MSE and high in LSE), and gene ontology analysis of the clusters demonstrated consistent cycle-phase-specific biological processes and molecular functions. Overall, the results demonstrated that endometrial samples obtained by two different sampling techniques (biopsy versus curettage of hysterectomy specimens) from healthy subjects including those with unknown histology can be classified by their molecular signatures in specific phases of the menstrual cycle with identical results using two independent analytical methods. These results underscore the potential to develop objective molecular diagnostics tool for profiling the human endometrium in health and disease.
Owing to the complexity of the endometrial tissue, and the possible dilution factor of genes present in specific compartments that are underrepresented such as the luminar or glandular epithelium, new strategies have been incorporated. Laser capture microdissection has been used to separate both stromal and epithelial compartments to analyse the differences in gene expression between them in mice (Niklaus and Pollard, 2006) and human (Yanaihara et al., 2005). It is a valuable approach to dissect out the genomic ‘weight’ between the two compartments. However, this approach requires a previous manipulation using frozen section that extends the time of analysis and can introduce errors because of the need of RNA amplification of the epithelial compartment.
|
Genomic analysis of the human endometrium during the window of receptivity in natural cycles |
|---|
In reproductive medicine, one of the most important challenges has been to be able to identify and diagnose the endometrial window of receptivity. In our laboratory, the initial studies were performed using macroarrays that contained 375 genes (Domínguez et al., 2003
). With the development of new microarrays, researchers started a particular race in this field and four studies on human endometrial gene expression were published in the period of 2002 and 2003 (Carson et al., 2002; Kao et al., 2002; Borthwick et al., 2003; Riesewijk et al., 2003). Recently, a further paper on this topic was published (Mirkin et al., 2005). Although all of these studies used the same technology, there are many differences in the experimental design and data analysis requires our attention, which include the day of the menstrual cycle for endometrial biopsy, the phases of the menstrual cycle compared, patient to patient genomic variation, number of endometrial biopsies and pooling or not pooling the isolated RNA (see Table II). Furthermore, different data analysis and statistical methods were employed for considering a gene regulated. Four of them established a minimal 2.0-fold increase to consider the gene regulation evident. However, our group considered a more stringent criterion of 3.0-fold increase, and the samples were obtained from the same patient at two different days of the menstrual cycle. This approach minimizes biological variability between samples (Riesewijk et al., 2003).
|
A partial comparison of microarray data from the five published studies is presented in the last of the five papers (Mirkin et al., 2005
). The differences indicated in study designs and methodology are reflected in the lack of a large list of consensus genes. Strikingly, only one gene, i.e. osteopontin, was consistently up-regulated in all the five studies. This structural protein, involved in cell adhesion, is present in several organs (Apparao et al., 2001). In the endometrium, it can be localized in the glands and secreted into the lumen (Johnson et al., 1999), and although it has assigned adhesive functions, its role in human embryonic implantation is still unknown (Johnson et al., 2003). There are several important molecules highlighted by their presence in four of five papers. Some of them are proteins previously identified in the endometrium with or without a described function. Genes involved in lipid metabolism (apolipoprotein D), immune response [decay accelerating factor for complement, serine or cysteine proteinase, interleukin (IL)-15], regulation of cell cycle (growth arrest and DNA-damage-inducible, alpha), ion binding (annexin IV) or enzymes with different functions in different tissues (monoamine oxidase A).
Interestingly, lipid metabolism is consistently controlled during endometrial development and specifically at the time of implantation. Kao’s work (Kao et al., 2002) underlines the presence of members of the Wnt family in the list of up-regulated genes. The marked up-regulation of Dickkopf-1 in all four studies is of particular interest. Dickkopf-1 inhibits Wnt signalling by binding to low-density lipid receptor-related protein (LRP5/6) (Mao et al., 2001). Wnt7A (–/–) null mice are infertile and have complete absence of uterine glands and a reduction in mesenchymally derived uterine stroma (Miller et al., 1998). The role of the Wnt family in human endometrium and implantation should be considered in future investigations.
These data offer the opportunity to develop an endometrial database (EDB) of genes expressed during the window of implantation (WOI) in the natural cycle. Nevertheless, it is important to acknowledge that the published data are complementary but not similar because of the differences indicated.
|
Genomic analysis of the human endometrium during controlled ovarian stimulated cycles |
|---|
Uterine receptivity is diminished during controlled ovarian stimulation (COS) used for IVF compared with natural cycles (Paulson et al., 1990
). The endometrium suffers a morphological advancement in the early luteal phase, which is demonstrated by histological techniques (Seif et al., 1992; Psychoyos, 1994; Kolb and Paulson, 1997; Kolibianakis et al., 2003), scanning electron microscopy (Nikas, 2000; Giudice, 2003), down-regulation of endometrial estrogen receptor and progesterone receptor (Develioglu et al., 1999) and biochemical changes in the endometrial fluid (Simón et al., 1996).
The impact of specific molecules in the development of endometrial receptivity has been reported from different perspectives, encompassing the genomic impact of progesterone on endometrial stromal cell development in vitro (Okada et al., 2003), IL-1ß (Rossi et al., 2005) or IL-11 (White et al., 2005). Recently, the in vitro effects of steroids on endometrial endothelial cells freshly isolated have been also reported (Krikun et al., 2005).
An attractive strategy is to investigate the genomic impact of identified COS protocols on endometrial receptivity during IVF treatment. In 2004, Mirkin and colleagues compared the gene expression profile in the peri-implantation endometrium in natural versus gonadotrophin-stimulated cycle using recombinant FSH (rFSH), with either GnRH agonist or GnRH antagonist, with or without progesterone supplementation of the luteal phase (Mirkin et al., 2004). Endometrial biopsies were collected in the previous natural cycle 8 days after the peak of LH (LH + 8) and 9 days after hCG administration (hCG + 9) in the next COS cycle. They concluded that although COS causes structural and functional changes compared with natural cycles, small changes were found when gene expression patterns were compared and that it may not have a major impact on endometrial receptivity. Only 18 genes/expressed sequence tags (ESTs) were found to be significantly different. They also concluded that significant changes were found when comparing cycles using GnRH agonist versus GnRH antagonist (13 genes significantly different) (Mirkin et al., 2004).
Two studies from our group have analysed the endometrium in COS cycles with different results. Our first paper assessed the endometrial impact of COS with a long protocol without progesterone supplementation. The endometrial profiles obtained at day hCG + 7 of COS were compared with those obtained at day LH + 7 of the previous natural cycle in the same patient; this design is important to avoid individual biological variability. In contrast to Mirkin et al. (2004), we found that >200 genes were dysregulated between COS and natural cycles (hCG + 7 versus LH + 7) (Horcajadas et al., 2005). These differences could be attributed to the following differences: the sample collection (samples from different patients versus the same patient) and the day of the endometrial collection (LH + 8/hCG + 9 versus LH + 7/hCG + 7).
Our second study evaluated the impact of standard and high doses of a GnRH antagonist (ganirelix) in stimulated cycles compared with GnRH agonist (buserelin); both protocols were supplemented with progesterone. All the groups were initiated with a fixed dose of rFSH, and endometrial biopsies were performed at hCG + 2 and hCG + 7 in COS cycles. We also added the backup of the endometrial collection at LH + 2 and LH + 7 from their previous natural cycle.
At hCG + 2, endometrial dating, estrogen and progesterone receptors, and pinopode appearance were comparable in all groups and the natural cycle. At hCG + 7, endometrial dating, steroid receptors and the presence of pinopodes were comparable in both GnRH antagonist groups and the natural cycles. In the long protocol, endometrial dating and pinopode expression suggested an arrested endometrial development compared with the other regimes. Gene expression profiles in the treatment cycles were largely comparable with that of the natural cycle at LH + 2. For WOI genes, expression patterns were closer to those in the natural cycle following standard- (50 genes dysregulated) or high-dose ganirelix (23 dysregulated) administration compared with buserelin administration (85 dysregulated) (Simón et al., 2005). To reflect clinical practice, we gave progesterone supplementation in the luteal phase in all three arms of the study. Under this homogenous condition, in each of the treatment groups, expression of about 100 genes was different from that in the natural cycle. This suggests that the endometrial gene dysregulation under COS is affected in a global manner and as a result, a different endometrial profile arises.
In conclusion, the endometrial genomic profile after daily treatment with standard- or high-dose GnRH antagonist in women undergoing COS mimics more closely the natural cycle compared with GnRH agonist. The clinical relevance of these findings remains to be elucidated. The number of dysregulated WOI genes found in these two studies (Horcajadas et al., 2005; Simón et al., 2005) are summarized in Table III.
|
|
Genomics of the eutopic/ectopic endometrium in endometriosis |
|---|
Endometriosis is a gynaecological disorder that occurs in
14% of women of reproductive age (Rice, 2002). It is a polygenic inherited disease of complex multi-factorial aetiology (Olive and Schwartz, 1993). The hypothesis is that the eutopic endometrium of women with endometriosis is different from that of non-endometriosis patients (Arici et al., 1996). In the pre-genomic era, when the investigation was limited by the inability to study more than one gene at a time, a number of genes related to galactose metabolism and detoxification enzymes (Zondervan et al., 2001), metalloproteases (Osteenet et al., 2003), aromatase (Bulun et al., 2004) and others were described. But now, in the genomic era, cDNA microarray technology has provided a great amount of useful information about the eutopic endometrium in endometriosis.
The first work that studied the gene expression profile of endometriosis using microarray technology was performed in 2002. Using a microarray that contained >4000 genes, researchers found eight genes up-regulated in endometriotic implants compared with uterine endometrium (Eyster et al., 2002). This pilot study was useful to demonstrate that the DNA microarray was an effective tool for the identification of differentially expressed genes between eutopic and ectopic endometria (Eyster et al., 2002).
In the same year, the effect of IL-1ß on endometriotic stromal cells was analysed using an array that contained 597 individual genes. IL-1ß is one of the most important cytokines involved in neovascularization and monocyte chemotaxis in endometriotic implants (Lebovic et al., 2000). The reported findings suggested that IL-1ß promotes growth of endometriotic lesions through inhibition of Tob-1, which was revealed to be clearly down-regulated in these experiments. It was also the first time that an alteration of cell-cycle gene expression was evidenced in cells derived from endometriotic implants (Lebovic et al., 2002).
The first analysis using high-density oligonucleotide microarrays compared the ectopic endometrium from women with endometriosis versus the endometrium from disease-free women. They investigated differentially regulated genes in the eutopic endometrium during the WOI (LH + 8 to LH + 10) from women with versus without endometriosis. The results showed the dysregulation of >200 genes in ectopic versus eutopic endometrium (Kao et al., 2003).
Another group has investigated ovarian endometriosis using Affymetrix technology (Arimoto et al., 2003). Matsuzaki etal. have analysed the genomic profile in deep endometriosis using laser capture microdissection and microarray (Matsuzaki et al., 2004, 2005).
In conclusion, all the mentioned studies have provided new insights into the endometriosis pathophysiology and have served to elaborate the gene expression profile of eutopic versus ectopic endometrium in this disease. There are genes currently known to be aberrantly expressed during the time of implantation in women with endometriosis, which may be candidates for initiation and establishment of the disease or for implantation-based infertility (Giudice and Kao, 2004). Among them, it is remarkable the presence of WOI genes such as glycodelin, leukaemia inhibitory factor (LIF) or glutathione peroxidase (Giudice and Kao, 2004).
|
Endometrial genes dysregulated in refractory conditions: RU486 and IUD |
|---|
An interesting approach to test the functionality of genes involved in human endometrial receptivity is to test the gene expression pattern of the endometrium in refractory conditions induced either pharmacologically or mechanically. The effect of RU486 on the gene expression profile in endometrial explants was investigated using a cDNA array containing
1000 verified gene targets which included genes known to be important in angiogenesis, apoptosis, cell signalling, extracellular matrix remodelling and cell-cycle regulation. Only 12 genes displayed significant changes in expression: six up-regulated and six down-regulated following RU386 treatment (Catalano et al., 2003). In this short list of genes, two important endometrial signalling pathways are included, i.e. JACK/STAT and JNK pathways. Using this approach, Sharkey et al. (2005) identified >100 transcripts as acutely regulated or dysregulated on LH + 7 by the administration of RU486 to women.
Our approach to study refractoriness was the investigation of the effect of an IUD on the endometrium in fertile patients. Over a period of 16 months, we analysed the gene expression profile of receptive versus refractory endometrium in the same patient (n = 5) induced by the presence of an inert IUD. The gene expression profile of the endometrium at LH + 7 before the insertion of the IUD, in the presence of IUD for 2 months and 1 year after the IUD removal was assessed. It has been demonstrated that the IUD prevents, at a molecular level, the normal transition to a receptive genomic status and identifies the genes that are responsible for the refractory status of the human endometrium. Surprisingly, the majority of genes that are dysregulated in the presence of the IUD remained dysregulated 2 months after IUD removal and recovered 1 year later (Horcajadas et al., 2006).
Undoubtedly, these approaches have generated valuable information about the physiology of human endometrium and provide a number of potential targets for the development of novel contraceptive drugs. The functional role of these molecules, involved in endometrial refractoriness, must be further analysed by means of functional analysis and in vivo models. It will be important to determine which of these is significantly involved in the refractoriness of the endometrium in the presence of IUD and can be taken into account as putative contraceptive targets.
|
Looking for the consensus genes in endometrial receptivity |
|---|
All the studies focused on endometrial receptivity have generated long lists of genes with known and unknown potential roles in this critical process. Analysing in detail the five works in natural cycles (Carson et al., 2002
; Kao et al., 2002; Borthwick et al., 2003; Riesewijk et al., 2003; Mirkin et al., 2005), we can observe that only one gene is consistently up-regulated, i.e. osteopontin. However, the number of genes increases considerably if we decrease the threshold to four or three matches in the five stages (Horcajadas et al., 2004a; Mirkin et al., 2005).
In order to limit a user-friendly list of key genes that could be relevant to endometrial receptivity, we decided to compare the results obtained in three different situations that included fertile conditions (natural cycle), subfertile (COS) and refractory conditions (IUD). We found 1399 WOI genes regulated in the human endometrium during the natural cycle (Riesewijk et al., 2003; Horcajadas et al., 2005). Of them, 342 genes showed to be increased or decreased more than twice during the WOI of COS cycles (Horcajadas et al., 2005). In the third study (Horcajadas et al., 2006), we found significant changes in 52 WOI genes in the endometrium in the presence of an inert IUD. After the comparison of the three studies, we found that the three works only shared 25 WOI genes (Figure 1). These genes are listed in Table IV. Interestingly, all of them showed to be regulated in one sense in the natural cycle, and on the contrary in COS and IUD (dysregulated). However, some of them (in bold) probably are not so important in the implantation process, because they did not recover their normal expression 1 year after the IUD removal, when the endometrial receptivity is considered to be normal. In Table IV, it is possible to find genes classically involved in endometrial receptivity such as glycodelin, LIF or 
-catenin. Furthermore, it has been demonstrated that some of them are also dysregulated in other situations such as endometriosis (Arimoto et al., 2003; Kao et al., 2003; Giudice and Kao, 2004) or endometrial cancer (Moreno-Bueno et al., 2003b).
|
|
|
Final considerations |
|---|
The significant histological, biological and physiological features that occur in the endometrium throughout the menstrual cycle are ultimately the result of changes that occur at the gene transcription level, together with the post-transcriptional modifications and epigenetic changes. Endometrial receptivity at the time of embryonic implantation is a crucial moment of the menstrual cycle with a fundamental relevance, and its understanding has been one of the main goals for researchers working in human reproduction.
Most of the laboratories have their favourite protein or molecule and have tried to elaborate its function in endometrial receptivity. But, at the moment, functional studies have not demonstrated the existence of a magic bullet for human endometrial receptivity. Probably, we will never be able to understand this complex process with the narrow focus of one gene, because it is the result of an equilibrated expression of many genes that conform to pathways and this concept must be kept in mind in future studies.
The studies analysed in this review have shown that endometrial receptivity is a very complex process, in which an uncountable number of genes are involved. However, these studies have also demonstrated that a limited number of candidates are always present and endometrial receptivity could be explained in terms of their modifications. Now is the time to learn about what the genomic era can add to our understanding of human endometrial receptivity. Future directions in endometrial receptivity studies will also require the complementarity with proteomics and functionomics. Although non-primate animal models have distinct advantages in economic and temporal cost, vast amounts of genetic information and the ability to be genetically modified, they remain inherently limited in their ability to elucidate the physiological mechanisms of human endometrial receptivity. However, studies on non-human primates have shown high fidelity to human implantation, suggesting their potential as models for investigation in endometrial receptivity, embryo implantation and early pregnancy.
Finally, the new analytical tools that appear periodically in the genomic field will provide to the researchers new comprehensive information to plan their experiments. Although at present there is no clear standard solution for microarray data storage and analysis software, there are many open-source, public domain and commercial solutions vying for a share of this evolving market. Most of the available products are in the very advanced phases of the software development process; however, new and improved versions of these are being released frequently to keep up with consumer expectations. We have created the EDB (http://www.endometrialdatabase.com) with the objective to facilitate the free exchange of information on the genomics of endometrial receptivity among basic and clinical endometriologists. We hope that the EDB became the perfect meeting point for the advancement of knowledge in this field worldwide. Researchers and bioinformatics experts should work together and learn from each other to improve both laboratory protocols and statistical tools to make possible a translational microarray technology.
|
References |
|---|
Ace CI and Okulicz WC (2004) Microarray profiling of progesterone-regulated endometrial genes during the rhesus monkey secretory phase. Reprod Biol Endocrinol 2,54.
Al-Shahrour F and Dopazo J (2005) In Azuaje F and Dopazo J (eds) Data Analysis and Visualization in Genomics and Proteomics. West Sussex, UK: Wiley, pp. 99–112.
Al-Shahrour F, Minguez P, Vaquerizas JM, Conde L and Dopazo J (2005) BABELOMICS: a suite of web tools for functional annotation and analysis of groups of genes in high-throughput experiments. Nucleic Acids Res 33,W460–W464.
Apparao KB, Murray MJ, Fritz MA, Meyer WR, Chambers AF, Truong PR and Lessey BA (2001) Osteopontin and its receptor alpha (v) beta (3) integrin are coexpressed in the human endometrium during the menstrual cycle but regulated differentially. J Clin Endocrinol Metab 86,4991–5000.
Arici A, Oral E, Bukulmez O, Duleba A, Olive DL and Jones EE (1996) The effect of endometriosis on implantation: results from the Yale University in vitro fertilization and embryo transfer program. Fertil Steril 65,603–607.[ISI][Medline]
Arimoto T, Katagiri T, Oda K, Tsunoda T, Yasugi T, Osuga Y, Yoshikawa H, Nishii O, Yano T, Taketani Y et al. (2003) Genome-wide cDNA microarray analysis of gene-expression profiles involved in ovarian endometriosis. Int J Oncol 22,551–560.[ISI][Medline]
Bammler T, Beyer RP, Bhattacharya S, Boorman GA, Boyles A, Bradford BU, Bumgarner RE, Bushel PR, Chaturvedi K, Choi D etal. (2005) Standardizing global gene expression analysis between laboratories and across platforms. Nat Methods 2,351–356.[CrossRef][ISI][Medline]
Barret JC and Kawasaki ES (2003) Microarrays: the use of oligonucleotides and cDNA for the analysis of gene expression. Drug Discov Today 8,134–141.[CrossRef][ISI][Medline]
Borthwick J, Charnock-Jones S, Tom BD, Hull ML, Teirney R, Phillips SC and Smith SK (2003) Determination of the transcript profile of human endometrium. Mol Hum Reprod 9,19–33.
Brar AK, Handwerger S, Kessler CA and Aronow BJ (2001) Gene induction and categorical reprogramming during in vitro human endometrial fibroblast decidualization. Physiol Genomics 7,135–148.
Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach J, Ansorge W, Ball CA, Causton HC et al. (2001) Minimum information about a microarray experiment (MIAME)—toward standards for microarray data. Nat Genet 29,365–371.[CrossRef][ISI][Medline]
Bulun SE, Fang Z, Imir G, Gurates B, Tamura M, Yilmaz B, Langoi D, Amin S, Yang S and Deb S (2004) Aromatase and endometriosis. Semin Reprod Med 22,45–50.[CrossRef][ISI][Medline]
Cao QJ, Belbin T, Socci N, Balan R, Prystowsky MB, Childs G and Jones JG (2004) Distinctive gene expression profiles by CDNA microarrays in endometrioid and serous carcinomas of the endometrium. Int J Gynecol Pathol 23,321–329.[ISI][Medline]
Carson D, Lagow E, Thathiah A, Al-Shami R, Farach-Carson MC, Vernon M, Yuan L, Fritz MA and Lessey B (2002) Changes in gene expression during the early to mid-luteal (receptive phase) transition in human endometrium detected by high-density microarray screening. Mol Hum Reprod 8,971–979.
Catalano RD, Yanaihara A, Evans AL, Rocha D, Prentice A, Saidi S, Print CG, Charnock-Jones DS, Sharkey AM and Smith SK (2003) The effect of RU486 on the gene expression profile in an endometrial explant model. Mol Hum Reprod 9,465–473.
Celera Genomics (2001) The sequence of the human genome. Science 291,1304–1351.
Cheon Y-P, Li Q, Xu X, DeMayo FJ, Bagchi IC and Bagchi MK (2002) A genomic approach to identify novel progesterone receptor regulated pathways in the uterus during implantation. Mol Endocrinol 16,2853–2871.
Churchill GA (2002) Fundamentals of experimental design for cDNA microarrays. Nat Genet 32,490–495.
Coutifaris C, Myers ER, Guzick DS, Diamond MP, Carson SA, Legro RS, McGovern PG, Schlaff WD, Carr BR, Steinkampf MP et al. (2004) Histological dating of timed endometrial biopsy tissue is not related to fertility status. Fertil Steril 82,1264–1272.[CrossRef][ISI][Medline]
Curtis-Hewitt S, Deroo BJ, Hansen K, Collins J, Grimson S, Afshari CA and Korach KS (2003) Estrogen receptor-dependent genomic responses in the uterus mirror the biphasic physiological response to estrogen. Mol Endocrinol 17,2070–2083.
Develioglu OH, Hsiu JG, Nikas G, Toner JP, Oehninger S and Jones HW Jr (1999) Endometrial estrogen and progesterone receptor and pinopode expression in stimulated cycles of oocyte donors. Fertil Steril 71,1040–1047.