Endocrinology Vol. 143, No. 6 2041-2044
Copyright © 2002 by The Endocrine Society
Perspective: The Ovarian Kaleidoscope Database—II. Functional Genomic Analysis of an Organ-Specific Database
Izhar Ben-Shlomo,
Ursula A. Vitt and
Aaron J. W. Hsueh
Division of Reproductive Biology, Department of Gynecology and Obstetrics, Stanford University School of Medicine, Stanford, California 94305-5317
Address all correspondence and requests for reprints to: Aaron J. W. Hsueh, Ph.D., Division of Reproductive Biology, Department of Gynecology and Obstetrics, Stanford University School of Medicine, 300 Pasteur Drive, Stanford, California 94305-5317. E-mail: . aaron.hsueh{at}stanford.edu
 |
Abstract
|
|---|
In the postgenomic era, it is now possible to investigate the function of all human genes to provide an integrated view of physiology and pathophysiology. An organ-based approach has been used to set up a database integrating existing text-based literature on individual ovarian genes and their sequence-based data in the GenBank. The Ovarian Kaleidoscope database (OKdb) has accumulated nearly one thousand individual gene pages that are searchable based on gene function, cellular localization, chromosomal position, ovarian cell type, ovarian function, mutant phenotypes, and other criteria. The present review exemplifies the use of this organ-based database in setting up gene pathway maps for DNA array analysis, identifying key gene networks essential for infertility phenotypes, comparing chromosomal synteny regions for finding candidate fertility genes, categorizing cell-specific and hormonally coregulated genes for promoter analysis, and documenting potential ligands and receptors in the paracrine regulation of follicular development. The present global analysis of gene function and relationships in an organ-specific manner provides a functional genomic paradigm for the future understanding of the physiology and pathophysiology of diverse organs.
|
Introduction
|
|---|
In recent years, there has been a dramatic increase in the amount of data on entire genomes of diverse species, profiles of gene transcription (the transcriptomes), and corresponding proteins (proteomes). However, the organizational principles of these massive data are sequence based or organism specific rather than organ specific. To understand the physiological roles of individual genes and their interactions in the coordinated regulation of tissue and organ functions, an organ-based approach is of value. The embryonic development of the ovary, the cyclic fluctuation of ovarian hormonal production throughout life, and the tight control of follicle maturation, ovulation, and luteinization are regulated by the large number of genes expressed in the ovary. A sorted and organized arrangement of the accumulating literature and gene sequence information would facilitate innovative research in ovarian physiology. The arrangement of this rapidly expanding data into a categorized searchable database has become prerequisite for the efficient analysis and understanding of whole organ systems. In September 1999, the OKdb (http://ovary.stanford.edu/) was launched as a unified online gateway to store, search, review, and update information about genes expressed in the ovary (1). Its infrastructure was designed to enable searches based on biological function, expression pattern, mutant phenotypes, cellular localization, and hormonal regulation of ovarian genes. The OKdb links to other online databases containing information on nucleotide and amino acid sequences, chromosomal localization, human and murine mutation phenotypes, and publications in PubMed that are relevant to individual genes.
The current number of gene entries in the OKdb is greater than 900 (2002) with approximately 7500 links to outside databases. Using the OKdb as an example of organ-specific functional genomic analysis, we herein describe bioinformatic approaches that were made possible due to the continuing accumulation of gene pages in the OKdb.
|
Construction of gene pathway maps
|
|---|
Protein function is usually accomplished by interactions with other proteins. Gene products in a given cell function in a concerted manner and belong to specific pathways for carrying out unique tasks. Although most DNA array experiments still focus on the analysis of transcripts with the largest increases or decreases in expression, new attempts have been made to investigate concerted regulation of the expression of genes in individual pathways. With the rapid development in microarray technology and the increasing numbers of genes that can be simultaneously probed, the GenMapp project has focused on the analysis and presentation of gene pathway maps (http://www.genmapp.org). This tool provides analogous graphic representation of the relative expression of selected groups of genes categorized by their known positions in physiological pathways. Thus, RNA transcripts for thousands of genes from defined cell populations can be probed for relative abundance based on prearranged maps representing functional pathways. A steroidogenic map is being incorporated into the OKdb, and each of the represented genes is linked to its gene page. One may now adapt the map to probe for the expression of steroidogenic genes in ovarian cells under various experimental conditions and view the results in direct connection with the existing text- and sequence-based knowledge of these genes.
Past research efforts focused on key genes within an intracellular pathway, whereas the map tool allows visualization of concomitant changes in the expression of all elements of a given pathway. Combining different pathway maps will eventually disclose a complex network of gene pathways to allow an integrated perspective on gene regulation in the cells and organs of interest. With a rapid expansion of data on genes in the ovulation pathway (2), integration of results from individual labs to construct a pathway map would be productive.
|
Ovarian "bottleneck" genes associated with infertility
|
|---|
A valuable approach to the study of infertility is the comparison of mutations of individual human and mouse genes associated with infertility phenotypes. The individual gene pages in the OKdb contain information on associated fertility phenotypes sorted by ovarian and nonovarian defects and by subfertility or infertility. If one searches for null mutations (under "mutation type") causing infertility ("infertile—ovarian defect," under "female fertility status") in mice (under "species"), 44 gene entries are found. The expression of these infertility genes in the oocyte and granulosa cells together with their cellular localization is presented in Fig. 1
. The theca cell genes are not presented because most publications emphasize granulosa cell studies.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 1. A graphic representation by cell type expression and cellular localization of infertility genes, based on phenotypes of mutant mice.
|
|
It is striking that all five oocyte infertility genes with known nuclear localization have been shown to be involved in DNA repair and mismatch correction, thus emphasizing the importance of this pathway(s) in regulating oocyte function and fertility. Although it is well known that ZP2 and ZP3 null mice are infertile, ZP1 null mice are only subfertile. Growth differentiation factor (GDF)-9 is an oocyte-secreted protein shown to regulate granulosa cell development. Although the receptor and downstream signaling pathway for GDF-9 is not clear, it is interesting to note that ALK-6 (a TGF-ß type 1 receptor family member) and SMAD 3 [an intracellular mediator of bone morphogenetic protein (BMP)/TGF-ß] null mice are also infertile. The majority of ovarian infertility genes expressed in granulosa cells are nuclear receptors and other transcription factors. The other prominent group of infertility genes in the granulosa cells is those associated with cell cycle regulation, cyclin D2, and the two cyclin associated genes CDN1B (cyclin-dependent kinase inhibitor 1B) and CDK4 (cyclin-dependent kinase 4).
Global analysis of these ovarian infertility genes and microarray analysis using a gene pathway approach can allow analysis of upstream and downstream component genes of these pathways. Analysis of DNA microarray studies, obtained under multiple experimental conditions or using tissues from knockout mice (3), could confirm and expand studies on "bottleneck" genes in key pathways that may underlie human ovarian pathologies.
|
Sorting out the cell-specific expression of genes
|
|---|
In the OKdb, one can retrieve genes based not only on their expression in a given cell type, but also in a cell type-specific manner. Presently, there are 324, 168, and 207 genes known to be expressed in the granulosa cell, theca cell, and oocyte, respectively. Whereas in the oocyte 53% of the listed genes are specific for the germ cell, in theca and granulosa cells only 8% and 29% of the listed genes are cell type specific, respectively (Fig. 2). It becomes clear that genes expressed in different ovarian somatic cell types overlap to a large extent, whereas the oocyte expresses a fundamentally different set of genes. In the case of cumulus cells, we found 63 expressed genes, of which only three were cumulus-specific: KE-6 (a 17-ß-hydroxysteroid dehydrogenase), Basigin, and TNF-
-induced protein 6, the latter likely to be important for cumulus expansion. The mismatch between the number of genes found only in one cell type and those defined as specific to that cell type, is due to the omission of genes expressed in luteal, stromal, epithelial, and other cell types in this analysis.
View larger version (71K):
[in this window]
[in a new window]
|
Figure 2. A chart representing the distribution of genes in the OKdb by their cellular expression in different ovarian cell types. The genes designated as specific are those expressed only in the oocyte or granulosa cells among all ovarian cell types.
|
|
The ability to search and define cell type-specific genes presents the unique opportunity to use the promoter regions of candidate genes for cell-specific knockout mice experiments (4). Recently, the Cre/loxP system has been implemented, leading to the deletion of a targeted gene only in the cell type of interest. Using this tissue-specific mutant mouse approach, it was shown that Bcl-x is not required for the maintenance of follicles and corpus luteum in the postnatal mouse ovary (5). In the future, comparison of ovarian genes with genes expressed in other tissues could provide leads to identify ovary-specific genes, mutations of which are likely to be associated with infertility phenotypes.
|
Promoter analysis of syn-expressed genes during follicular development, ovulation, and luteinization
|
|---|
Genes are expressed in a development-, tissue-, and regulation-specific manner. The promoter sequences of all genes contain multiple transcription factor binding sites and the expression of individual genes is dependent on a combinatorial use of transcription factors at multiple promoter elements. As shown in S. cerevisiae, combinatorial analysis of promoter elements could improve our understanding of gene regulation (6). A critical issue for all large-scale computational approaches to gene regulation is, however, the experimental validation of the physiological significance of the predicted promoter elements (7). The detailed spatio-temporal expression of ovarian genes, as documented in the OKdb, can serve as the basis for future identification and verification of common regulatory mechanisms of gene expression, including consensus DNA elements and transcriptional factors. The available data in the OKdb can be used to construct microarray queries based on the existing knowledge of ovarian gene expression such as genes induced by gonadotropins and growth factors. This approach could facilitate the confirmation and identification of syn-expressed genes for future promoter analysis using bioinformatic methods.
|
Chromosomal localization of genes: identification of mutant genes with unique phenotypes
|
|---|
Completion of the human genome project not only provides information about the sequence of individual genes but specifies their chromosomal location. Extensive genetic analysis has accumulated a vast amount of literature on the chromosomal localization of mutant genes associated with unique phenotypes. The current structure of the OKdb includes a direct link for each listed gene to its chromosomal localization in human as well as a human chromosomal map of ovarian genes. Whereas in phylogenetically distant animals gene localizations on chromosomes are unrelated, in phylogenetically close animals, such as humans, mice, and sheep, the conservation of gene order in chromosome segments (synteny) constitutes a powerful research approach.
Fecundity is a unique trait that has been studied extensively in humans and farm animals. The GDF-9 and GDF-9B (also known as BMP-15) signaling pathways play an essential role in follicle development. Loss-of-function mutations in GDF-9 null mice (8) and GDF-9B mutant Inverdale sheep (9) are associated with an arrest of initial follicle recruitment (10). Paradoxically, the heterozygous female Inverdale sheep exhibited an increase in the number of offspring, presumably due to an increase in the number of ovulated oocytes. Interestingly, the Booroola FecB mutation in sheep, causing increased fecundity, is phenotypically identical to the heterozygous Inverdale sheep. Cosegregation studies localized the Fec B mutation to chromosome 6 in sheep (11), syntenic to the human chromosome 4q (12). Based on the assumption that similar phenotypes are likely to be found for ligand/receptor pairs, one can search for receptors located at human 4q in the OKdb. Of interest, ALK-6, a family member of the TGF-ß receptor family, is localized in this region and represents a candidate receptor for GDF-9 or GDF-9B. One can predict that ALK-6 is a candidate gene for the Booroola sheep mutation. Indeed, a mutation in the intracellular kinase domain of ALK-6 was found in the Booroola sheep (13, 14, 15).
Future use of chromosomal locations may reveal candidate ovarian genes responsible for ovarian pathologies. The polycystic ovary (PCO) syndrome, exhibiting hyperandrogenic anovulation, has long intrigued reproductive endocrinologists. One of the promising approaches to the study of this pathology is the cosegregation of genes within affected families. This has led to the identification of several candidate PCO genes, including CYP11, insulin receptor, and follistatin (16, 17). However, the genetic approach provides only correlative information and usually maps the genes to a large chromosomal segment. Studies on the chromosomal location of all ovarian genes could facilitate the elucidation of ovarian genes in the pathogenesis of PCO syndrome.
As the chromosomal localization of most mouse genes are known, a portal for synteny searches has been established by NCBI http://www.ncbi.nlm.nih.gov/Homology/. With the well-developed technology of gene engineering in mouse, future comparisons of human and mouse phenotypes and the elucidation of underlying genes based on chromosomal synteny approaches are of value.
|
Identification of ligand-receptor pairs for paracrine regulation
|
|---|
It is becoming clear that all organs are under paracrine control and autocrine and paracrine regulators for any organ outnumber the endocrine regulators. For example, follicular development is driven by the signaling cascade initiated by FSH and LH but is fine-tuned by numerous paracrine and juxtacrine interactions (18). The ovary, like all other organs, is endowed with a multitude of paracrine networks of ligands and receptors. Although earlier studies clearly documented the importance of insulin-like growth factor-I (19), interleukin-1 (20), TGF-ß/BMPs (21), and kit ligand (22) in ovarian physiology, few attempts have been made to obtain a comprehensive understanding of ovarian paracrine networks.
A prerequisite for paracrine regulation is the expression of both ligands and their cognate receptors in the same organ. The OKdb accumulates literature- and sequence-based information (including ESTs) on gene expression in the ovary. By searching the OKdb, one can identify potential paracrine interacting pairs. The most striking of these are the ovarian expression of one Notch receptor family member and one Notch ligand, four Toll receptor family members and three Toll ligands as well as two guanylyl cyclase family receptors with two ligands. Armed with this information, ovarian researchers can further investigate the spatial and temporal relationships of the expression of these genes to uncover their physiological roles. A large-scale study using this approach has been conducted recently on malignant tumors to define the presence of paracrine ligand-receptor pairs likely involved in tumorigenesis (23).
In summary, an organ-specific database such as the OKdb has a unique role in functional genomics. It brings together text- and sequence-based data for the compilation of information on the expression and function of individual genes in the organ of interest, thus allowing an integrated global perspective on gene pathways and networks. Implementation of an organ-based bioinformatic paradigm is anticipated to be productive for postgenomic research.
|
Acknowledgments
|
|---|
|
Footnotes
|
|---|
This work was supported by the NICHHD/NIH through Cooperative Agreement U54-HD-31398 as part of the Specialized Cooperative Centers Program in Reproduction Research. I.B.-S. was supported by the Feldman Foundation; on leave from HaEmek Medical Center (Afula, Israel) and the Rappaport Faculty of Medicine, Technion-Israel Institute of Technology (Haifa, Israel).
Abbreviations: BMP, Bone morphogenetic protein; GDF, growth differentiation factor; OKdb, Ovarian Kaleidoscope database; PCO, polycystic ovary.
Received February 12, 2002.
Accepted for publication February 20, 2002.
|
References
|
|---|
-
Leo CP, Vitt UA, Hsueh AJ 2000 The Ovarian Kaleidoscope database: an online resource for the ovarian research community. Endocrinology 141:3052–3054[Abstract/Free Full Text]
-
Richards JS, Russell DL, Ochsner S, Espey LL 2002 OVULATION: new dimensions and new regulators of the inflammatory-like response. Annu Rev Physiol 64:69–92[CrossRef][Medline]
-
Elvin JA, Matzuk MM 1998 Mouse models of ovarian failure. Rev Reprod 3:183–195[Abstract]
-
Sauer B 1998 Inducible gene targeting in mice using the Cre/lox system. Methods 14:381–392[CrossRef][Medline]
-
Riedlinger G, Okagaki R, Wagner KU, Rucker IE, Oka T, Miyoshi K, Flaws JA, Hennighausen L 2002 Bcl-x is not required for maintenance of follicles and corpus luteum in the postnatal mouse ovary. Biol Reprod 66:438–444[Abstract/Free Full Text]
-
Pilpel Y, Sudarsanam P, Church GM 2001 Identifying regulatory networks by combinatorial analysis of promoter elements. Nat Genet 29:153–159[CrossRef][Medline]
-
Michelson AM 2002 Deciphering genetic regulatory codes: a challenge for functional genomics. Proc Natl Acad Sci USA 99:546–548[Free Full Text]
-
Dong J, Albertini DF, Nishimori K, Kumar TR, Lu N, Matzuk MM 1996 Growth differentiation factor-9 is required during early ovarian folliculogenesis. Nature 383:531–535[CrossRef][Medline]
-
Galloway SM, McNatty KP, Cambridge LM, Laitinen MP, Juengel JL, Jokiranta TS, McLaren RJ, Luiro K, Dodds KG, Montgomery GW, Beattie AE, Davis GH, Ritvos O 2000 Mutations in an oocyte-derived growth factor gene (BMP15) cause increased ovulation rate and infertility in a dosage-sensitive manner. Nat Genet 25:279–283[CrossRef][Medline]
-
McGee EA, Hsueh AJ 2000 Initial and cyclic recruitment of ovarian follicles. Endocr Rev 21:200–214[Abstract/Free Full Text]
-
Montgomery GW, Lord EA, Penty JM, Dodds KG, Broad TE, Cambridge L, Sunden SL, Stone RT, Crawford AM 1994 The Booroola fecundity (FecB) gene maps to sheep chromosome 6. Genomics 22:148–153[CrossRef][Medline]
-
Montgomery GW, Crawford AM, Penty JM, Dodds KG, Ede AJ, Henry HM, Pierson CA, Lord EA, Galloway SM, Schmack AE, Sise JA, Swarbrick PA, Hanrahan V, Buchanan FC, Hill DF 1993 The ovine Booroola fecundity gene (FecB) is linked to markers from a region of human chromosome 4q. Nat Genet 4:410–414[CrossRef][Medline]
-
Wilson T, Wu XY, Juengel JL, Ross IK, Lumsden JM, Lord EA, Dodds KG, Walling GA, McEwan JC, O’Connell AR, McNatty KP, Montgomery GW 2001 Highly prolific Booroola sheep have a mutation in the intracellular kinase domain of bone morphogenetic protein IB receptor (ALK-6) that is expressed in both oocytes and granulosa cells. Biol Reprod 64:1225–1235[Abstract/Free Full Text]
-
Mulsant P, Lecerf F, Fabre S, Schibler L, Monget P, Lanneluc I, Pisselet C, Riquet J, Monniaux D, Callebaut I, Cribiu E, Thimonier J, Teyssier J, Bodin L, Cognie Y, Chitour N, Elsen JM 2001 Mutation in bone morphogenetic protein receptor-IB is associated with increased ovulation rate in Booroola Merino ewes. Proc Natl Acad Sci USA 98:5104–5109[Abstract/Free Full Text]
-
Souza CJ, MacDougall C, Campbell BK, McNeilly AS, Baird DT 2001 The Booroola (FecB) phenotype is associated with a mutation in the bone morphogenetic receptor type 1 B (BMPR1B) gene. J Endocrinol 169:R1–R6
-
Urbanek M, Legro RS, Driscoll DA, Azziz R, Ehrmann DA, Norman RJ, Strauss 3rd JF, Spielman RS, Dunaif A 1999 Thirty-seven candidate genes for polycystic ovary syndrome: strongest evidence for linkage is with follistatin. Proc Natl Acad Sci USA 96:8573–8578[Abstract/Free Full Text]
-
Franks S, Gharani N, McCarthy M 2001 Candidate genes in polycystic ovary syndrome. Hum Reprod Update 7:405–410[Abstract/Free Full Text]
-
Hsueh AJ, Adashi EY, Jones PB, Welsh Jr TH 1984 Hormonal regulation of the differentiation of cultured ovarian granulosa cells. Endocr Rev 5:76–127[Abstract/Free Full Text]
-
Adashi EY 1998 The IGF family and folliculogenesis. J Reprod Immunol 39:13–19[CrossRef][Medline]
-
Adashi EY 1998 The potential role of interleukin-1 in the ovulatory process: an evolving hypothesis. Mol Cell Endocrinol 140:77–81[CrossRef][Medline]
-
Elvin JA, Yan C, Matzuk MM 2000 Oocyte-expressed TGF-ß superfamily members in female fertility. Mol Cell Endocrinol 159:1–5[CrossRef][Medline]
-
Driancourt MA, Reynaud K, Cortvrindt R, Smitz J 2000 Roles of KIT and KIT LIGAND in ovarian function. Rev Reprod 5:143–152[Abstract]
-
Graeber TG, Eisenberg D 2001 Bioinformatic identification of potential autocrine signaling loops in cancers from gene expression profiles. Nat Genet 29:295–300[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
A Bonnet, R Dalbies-Tran, and M A Sirard
Opportunities and challenges in applying genomics to the study of oogenesis and folliculogenesis in farm animals
Reproduction,
February 1, 2008;
135(2):
119 - 128.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. R. Wood, D. A. Dumesic, D. H. Abbott, and J. F. Strauss III
Molecular Abnormalities in Oocytes from Women with Polycystic Ovary Syndrome Revealed by Microarray Analysis
J. Clin. Endocrinol. Metab.,
February 1, 2007;
92(2):
705 - 713.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. M. A. Tullet, V. Pocock, J. H. Steel, R. White, S. Milligan, and M. G. Parker
Multiple Signaling Defects in the Absence of RIP140 Impair Both Cumulus Expansion and Follicle Rupture
Endocrinology,
September 1, 2005;
146(9):
4127 - 4137.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
E. Jansen, J. S. E. Laven, H. B. R. Dommerholt, J. Polman, C. van Rijt, C. van den Hurk, J. Westland, S. Mosselman, and B. C. J. M. Fauser
Abnormal Gene Expression Profiles in Human Ovaries from Polycystic Ovary Syndrome Patients
Mol. Endocrinol.,
December 1, 2004;
18(12):
3050 - 3063.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
N. Massin, A. Gougeon, G. Meduri, E. Thibaud, K. Laborde, C. Matuchansky, E. Constancis, M.C. Vacher-Lavenu, B. Paniel, J.R. Zorn, et al.
Significance of ovarian histology in the management of patients presenting a premature ovarian failure
Hum. Reprod.,
November 1, 2004;
19(11):
2555 - 2560.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. R. Wood, V. L. Nelson, C. Ho, E. Jansen, C. Y. Wang, M. Urbanek, J. M. McAllister, S. Mosselman, and J. F. Strauss III
The Molecular Phenotype of Polycystic Ovary Syndrome (PCOS) Theca Cells and New Candidate PCOS Genes Defined by Microarray Analysis
J. Biol. Chem.,
July 11, 2003;
278(29):
26380 - 26390.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
