Skip Navigation


Human Molecular Genetics Advance Access originally published online on January 6, 2006
Human Molecular Genetics 2006 15(3):417-431; doi:10.1093/hmg/ddi463
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Material
Right arrow All Versions of this Article:
15/3/417    most recent
ddi463v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (20)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Beverdam, A.
Right arrow Articles by Koopman, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Beverdam, A.
Right arrow Articles by Koopman, P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2005. Published by Oxford University Press. All rights reserved.
For Permissions, please email: journals.permissions@oxfordjournals.org

Expression profiling of purified mouse gonadal somatic cells during the critical time window of sex determination reveals novel candidate genes for human sexual dysgenesis syndromes

Annemiek Beverdam1 and Peter Koopman1,2,*

1Division of Genetics and Developmental Biology and 2ARC Centre of Excellence in Biotechnology and Development, Institute for Molecular Bioscience, The University of Queensland, Brisbane, Qld 4072, Australia

* To whom correspondence should be addressed. Tel: +61 733462059; Fax: +61 733462101; Email: p.koopman{at}imb.uq.edu.au

Received October 17, 2005; Accepted December 11, 2005


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Despite the identification of SRY as the testis-determining gene in mammals, the genetic interactions controlling the earliest steps of male sex determination remain poorly understood. In particular, the molecular lesions underlying a high proportion of human XY gonadal dysgenesis, XX maleness and XX true hermaphroditism remain undiscovered. A number of screens have identified candidate genes whose expression is modulated during testis or ovary differentiation in mice, but these screens have used whole gonads, consisting of multiple cell types, or stages of gonadal development well beyond the time of sex determination. We describe here a novel reporter mouse line that expresses enhanced green fluorescent protein under the control of an Sf1 promoter fragment, marking Sertoli and granulosa cell precursors during the critical period of sex determination. These cells were purified from gonads of male and female transgenic embryos at 10.5 dpc (shortly after Sry transcription is activated) and 11.5 dpc (when Sox9 transcription begins), and their transcriptomes analysed using Affymetrix genome arrays. We identified 266 genes, including Dhh, Fgf9 and Ptgds, that were upregulated and 50 genes that were downregulated in 11.5 dpc male somatic gonad cells only, and 242 genes, including Fst, that were upregulated in 11.5 dpc female somatic gonad cells only. The majority of these genes are novel genes that lack identifiable homology, and several human orthologues were found to map to chromosomal loci implicated in disorders of sexual development. These genes represent an important resource with which to piece together the earliest steps of sex determination and gonad development, and provide new candidates for mutation searching in human sexual dysgenesis syndromes.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Disorders of sexual development encompass a wide spectrum of conditions ranging in severity from hypospadias (inappropriate location of the urethral opening in males) to intersex, premature ovarian failure and complete sex reversal syndromes. Although relatively mild and easily treated, hypospadias is regarded as the most frequent complication in newborn males. In contrast, intersex and sex reversal syndromes are far less common, but often result in infertility, genital abnormalities, gender mis-assignment and long-term psychological trauma. At the root of these disorders is often the complex network of gene regulation underpinning the correct development of the testes or ovaries from their common embryonic precursors, the genital ridges.

In mice, the genital ridges develop around 9.5 days post coitum (dpc) on the ventral surface of the mesonephroi. They are populated by germ cells that migrate from the hindgut from around 10.5 dpc. In XY embryos, the Y-linked gene Sry is transiently activated around 10 dpc in the supporting cell lineage, which also expresses the orphan nuclear receptor gene Sf1 (1Go–6Go). Sry induces the differentiation of pre-Sertoli cells, marked by the expression of a second male sex-determining gene, Sox9, by 11.5 dpc (7Go,8Go). Pre-Sertoli cells are considered to be the organizers of testis differentiation, producing key signalling molecules that influence the differentiation of other testicular cell types and their organization into the characteristic testicular histology.

Among the other cell types in the developing testis are the steroidogenic Leydig cells. Precursors of this cell type are thought to ingress from the epithelium at the coelomic surface of the gonad (9Go,10Go), or from the mesonephros (11Go,12Go), and initially do not express Sf1. Shortly after onset of their differentiation at around 12.0 dpc, they migrate into the gonads where they first activate Sf1 expression (9Go) and, later, produce testosterone which induces development of the male reproductive tract. In addition, vascular and myoid cells are recruited from the mesonephros (13Go–15Go). By 12.5 dpc, male gonads have acquired the basic organization of testes: sex cords have formed, consisting of clusters of germ cells surrounded by Sertoli cells, basement membrane and a flattened layer of peritubular myoid cells, whereas Leydig, vascular and other cells occupy the interstitial space between the sex cords.

In contrast, female gonads remain in a poorly defined, ‘undifferentiated’ state during this period and do not display clear female-specific morphology until several days later in the mouse. However, follicular organization is pre-figured by the early, female-specific upregulation of several genes such as Foxl2, Fst, Dax1 and Wnt4 (16Go–19Go), but it is not yet clear what cell populations exist in the early ovary or how these are distinguishable by expression of marker genes (for review, see 20Go). By the time of birth, primordial follicles have developed, consisting of post-meiotic oocytes, surrounded and supported by granulosa cells (21Go–23Go). By analogy with the role of pre-Sertoli cells in the testis, it is assumed that pre-granulosa cells play an important role in orchestrating ovarian differentiation, but at present the genes and molecular events driving the differentiation of ovarian cell types and their organization into functional ovaries remain almost completely unknown.

It is clear that much remains to be learned about the genetic networks that underlie the first steps of sex determination and gonad differentiation and that defects in these processes might underlie the large proportion of human gonadal dysgenesis, intersex and sex reversal syndromes that remain unexplained (24Go).

A number of groups have undertaken expression screens to identify novel genes that may fill the gaps in our knowledge of sex determination and the aetiology of intersex disorders. Wertz and Herrmann (25Go) undertook a high-throughput whole mount in situ hybridization screen using probes generated from an 11.5 dpc mouse genital ridge cDNA library. Several laboratories have employed subtraction screens to identify genes differentially expressed in male and female mouse gonads at 12.0–13.5 dpc (26Go–29Go) and between 12.0 and 12.5 dpc mouse Sertoli cells (30Go). Others have used microarrays to determine the expression profiles of whole male and female embryonic mouse gonads and carried out in silico subtractions to identify genes with dimorphic expression patterns in male and female gonads (31Go,32Go). Although each screen has identified known and novel genes with sexually dimorphic expression patterns in gonads of 11.5 dpc and older, none of these screens has addressed the critical time window for male sex determination between 10.5 and 11.5 dpc. Moreover, most of these screens have been based on material extracted from whole gonads that consist of complex mixture of cell types. This approach is likely to obscure the expression of genes that regulate and characterize the supporting cells of the developing gonads and hence may play the key roles in sex determination and gonadal differentiation.

In this study, we generated, characterized and exploited a transgenic reporter mouse line, in which enhanced green fluorescent protein (EGFP) is expressed under control of a 674 bp Sf1 promoter fragment exclusively in somatic cells of male and female gonads (33Go). This reporter mouse line enabled us to determine the expression profiles of somatic gonadal cells during the critical time window of male sex determination. Although some genes with known roles during early gonad development were identified in our screen, attesting to the validity of the tools and approaches used, many were novel genes devoid of structural, molecular or functional annotation. In addition, we found that a number of the human orthologues of these genes map to human chromosomal loci known to be affected in human sex disorders. This study, therefore, provides important new tools that can be used to unravel the genetic basis of the earliest steps of sex determination and early male and female gonad development, and may yield new insights into the aetiology of human sexual dysgenesis syndromes.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
pSf1–EGFP transgenic embryos express EGFP in somatic cells of the developing gonads
To obtain more insight into the genetic networks controlling the earliest stages of male and female sex determination and gonad development, we first set out to generate transgenic mice that expressed EGFP in the supporting cell lineage of both male and female gonad cells. Sf1 is expressed in somatic gonad cells of male and female gonads from 9.5 dpc onwards (34Go). It has recently been shown that LHX9 and WT1 binding sites within a 674 bp Sf1 promoter fragment (–590 to +85) are responsible for directing expression of Sf1 into somatic gonad cells (33Go). We, therefore, used this promoter fragment to drive expression of EGFP in transgenic mice.

Two transgenic lines were established that expressed EGFP in embryonic gonads at different levels (Fig. 1B–E and data not shown). Expression was also observed in non-gonadal sites that differed between the two lines. All further analyses were carried out with the transgenic mouse line showing the higher levels of EGFP expression. These mice expressed EGFP ectopically in dorsal root ganglia, nasal processes and limb mesenchyme at 10.5 dpc (Fig. 1A) and bone condensations in cranial and outgrowing limb buds, ribs and vertebrae at 13.5 dpc (data not shown).


Figure 4631
View larger version (113K):
[in this window]
[in a new window]
  		Figure 1. EGFP is expressed in XX and XY gonads of pSf1–EGFP transgenic embryos. (A) Whole 10.5 dpc pSf1–EGFP transgenic embryo showing expression of EGFP in the nasal processes (NP), proximal limb buds (L), dorsal root ganglia (DRG) and genital ridges (GR). (B) Dissected genital ridges of 10.5 dpc pSf1–EGFP transgenic embryo showing EGFP expression in cells scattered throughout the genital ridges (outlined). (C) Dissected XY gonad of a 13.5 dpc pSf1–EGFP transgenic embryo showing EGFP expression in the testis cords (one of which is shown outlined), but not in the mesonephros (M). (D) Dissected XX gonad of a 13.5 dpc pSf1–EGFP transgenic embryo showing EGFP expression in a subset of cells throughout the gonad, but not in the mesonephros (M). (E) Dissected right half of the urogenital system of an XY 13.5 dpc pSf1–EGFP transgenic embryo, showing expression within the testis cords of the gonad, but not in the kidney (K) or in the adrenal anlage (A).

 
We detected EGFP expression in male and female genital ridges as early as 10.5 dpc, persisting to 16.5 dpc, the latest stage tested, but not in the mesonephroi (Fig. 1B–D). However, no EGFP expression was observed in the anlage of the adrenal cortex, the developing pituitary primordium or in the developing brain or spleen, all tissues that normally express Sf1 (Fig. 1E and data not shown). This restricted expression precluded contamination of the flow-sorted EGFP-positive cells used for profiling with nascent adrenal cells.

To characterize the expression of the marker transgene in detail, we sectioned male and female gonads at a range of stages and performed immunostainings with antisera directed against the germ-cell-specific protein Mouse Vasa Homologue (MVH) (35Go), the extracellular matrix protein laminin, and SF1 itself. In 11.5 dpc male and female gonads, we observed EGFP expression exclusively in cells that did not express MVH (Fig. 2A and B and data not shown). In addition, most if not all SF1-expressing cells were found to co-express EGFP (Fig. 2C and D and data not shown), confirming that the EGFP-expressing cells are somatic gonad cells and that transgene expression marked the cell population that normally expresses Sf1.


Figure 4632
View larger version (82K):
[in this window]
[in a new window]
  		Figure 2. EGFP is expressed in SF1-positive somatic gonad cells. (A–D) Sections of an 11.5 dpc XY gonad stained with antibodies directed against MVH (A, B) and SF1 (C, D) at 20x (A, C) and 63x (B, D) magnification. (E and F) Sections of a 12.5 dpc XY gonad stained with antibodies directed against Laminin (LAM) at 20x (E) and 63x (F) magnification. (G–J) Sections of a 14.5 dpc XY gonad stained with antibodies directed against MVH (G, H) and LAM (I, J) at 20x (G, I) and 63x (H, J) magnification. In (I), one testis cord is marked with TC. Arrows in (J) point to Sertoli (S) and Leydig (L) cells. The arrowhead in (J) points to a cell that may be a peritubular myoid cell. Scale bars, 100 µm.

 
In 12.5 dpc male gonads, strongly EGFP-positive cells were found within the testis cords outlined by anti-laminin staining, suggesting that transgene expression marked the Sertoli cell lineage at this and earlier stages (Fig. 2E and F). At this stage, cells expressing EGFP more weakly were detected in the interstitium (Fig. 2E and F), consistent with the reported upregulation of Sf1 in the Leydig cells at this stage (9Go,36Go). In 14.5 dpc male gonads, we detected cells expressing EGFP, but not MVH (Fig. 2G and H), and co-staining with anti-laminin antibody showed that EGFP is expressed both by Sertoli cells within the cords (labelled S in Fig. 2J) and by Leydig cells in the interstitial space (labelled L in Fig. 2J) at this stage. We observed some flattened EGFP-positive cells circumscribing the testis cords (arrowhead in Fig. 2J). It remains unclear whether these are peritubular myoid cells, as no reliable marker exists for these cells (37Go). In 12.5 and 14.5 dpc female gonads, we detected EGFP expression exclusively in SF1-positive and MVH-negative gonad cells (data not shown).

In summary, pSf1–EGFP transgenic embryos were found to express EGFP exclusively in somatic gonad cells, including supporting cells (Sertoli and granulosa cell precursors) at 10.5 and 11.5 dpc, and, later, in Sertoli and Leydig cells, but not in adrenal cells. This transgenic mouse line is, therefore, ideal for the purpose of transcriptional-profiling early-supporting-cell-precursor cells with a view to identifying novel candidate sex-determining genes.

Marker analysis of sorted cells
To affirm the conclusions of our immunofluorescence assays and to characterize the cell populations that could be purified by flow sorting, we next dissected 11.5 and 13.5 dpc transgenic gonads, including the mesonephroi, from male and female embryos, dissociated them into single-cell suspensions and collected EGFP-positive and -negative cell fractions using fluorescence-activated cell sorting (FACS). RNA was extracted from the sorted cells and used for quantitative real-time RT–PCR analysis of expression of known somatic and germ cell markers.

We found that somatic cell markers Sf1, Sox9 and Sry were expressed robustly in EGFP-positive male gonadal ridge cell populations at 11.5 dpc and at negligible levels in the corresponding EGFP-negative cell populations (Fig. 3 and data not shown). Similarly, somatic cell markers Sf1 and Dax1 were expressed almost exclusively in the female EGFP-positive gonadal ridge cell fraction at 11.5 dpc (Fig. 3). This analysis indicates that the EGFP-positive fraction expresses markers of the supporting cell lineage and that the EGFP-negative fraction is not contaminated with these cells.


Figure 4633
View larger version (22K):
[in this window]
[in a new window]
  		Figure 3. EGFP-positive cells express somatic cell markers, whereas EGFP-negative cells express germ cell markers. Results of quantitative PCR experiments carried out on RNA extracted from 11.5 dpc XY gonad cells (top row), 11.5 dpc XX gonad cells (middle row) and 13.5 dpc XY gonad cells (bottom row). Dark bars represent EGFP-positive RNA samples, whereas white bars represent EGFP-negative RNA samples. The Y-axes display relative fold expression levels compared with 18S rRNA. The error bars represent minimal and maximal ranges of gene expression levels in four replicates when compared with 18S rRNA.

 
In 13.5 dpc male transgenic gonads, we found significantly higher expression levels of Sf1, Sox9 and Cyp11a1 in EGFP-positive cells than in EGFP-negative cells, confirming the presence of both Sertoli and Leydig cells in the EGFP-positive male gonad cell population at this later stage (Fig. 3), as expected from our immunofluorescence data.

Oct4, a marker for germ cells, was found to be expressed in EGFP-negative male and female cell populations at 11.5 dpc (Fig. 3), as expected. Oct4 expression levels were significantly higher in EGFP-negative cells than in the corresponding EGFP-positive populations (Fig. 3), indicating that the EGFP-positive cell population was enriched for somatic cells. The gene expression profiles of EGFP-positive cell populations are, therefore, expected to reflect predominantly the transcriptional activities of the somatic cell population, in line with the aims of this study.

Expression profiling of GFP-positive populations of 10.5 and 11.5 dpc male and female gonads
Two time points, 10.5 and 11.5 dpc, were analysed to identify transcriptional changes in somatic cells during the critical time window of male and female sex determination. At 10.5 dpc, Sry expression is first activated in supporting cells of male gonads, whereas at 11.5 dpc, Sox9 expression is first detected in the same cells (7Go). Genital ridges, including mesonephroi and dorsal mesentery, were dissected from male (n=21) and female (n=16) transgenic embryos at 10.5 dpc and from male (n=31) and female (n=48) transgenic embryos at 11.5 dpc (Table 1). The genital ridges were dissociated to single-cell suspensions, and EGFP-positive and -negative cell populations were separated using FACS, quantified and collected.


View this table:
[in this window]
[in a new window]
  		Table 1. FACS statistics
 
According to these analyses, about 410 cells per male and 320 cells per female genital ridge were found to express EGFP at 10.5 dpc (Table 1). At 11.5 dpc, however, there were significantly more EGFP-positive cells present in male than in female genital ridges (4200 in male versus 1900 in female; Table 1). These observations are most likely due to the male-specific increase in proliferation that occurs in gonadal somatic cells immediately following activation of Sry expression (9Go).

RNA was extracted from the flow-sorted cell samples. Because of the limited amount of nucleic acid available from this small number of cells, sample RNA was linearly amplified in two consecutive rounds, labelled with biotin and fragmented. Amplified RNA derived from EGFP-positive cells of 10.5 and 11.5 dpc male and female gonads was hybridized in triplicate to GeneChip® Mouse Genome 430 2.0 Arrays (Affymetrix). The data sets were normalized and further analysed in GeneSpring. Subtractions were carried out between the EGFP-positive RNA samples to identify genes the expression of which differed at least 2-fold between samples, with a statistical confidence level of greater than 95% (P<0.05). All microarray data resulting from these experiments are publicly available on the Gene Expression Omnibus (GEO) Web site (www.ncbi.nlm.nih.gov/geo/) and can be accessed with accession number GSE3463 [NCBI GEO] .

Genes upregulated in male somatic cells only
Genes exclusively upregulated in 11.5 dpc male somatic gonad cells (compared with male somatic gonad cells at 10.5 dpc and female somatic gonad cells at 11.5 dpc) are most likely those that direct the earliest processes of male sex determination and early gonad development. In particular, these are likely to comprise direct SRY target genes, and genes that induce Sox9 expression, and may include yet unidentified male sex-determining genes. We identified 1873 genes that were expressed at higher levels in EGFP-positive gonad cell populations of 11.5 dpc versus 10.5 dpc male embryos (>2-fold, P<.0.05; Fig. 4A). Moreover, 796 transcripts were expressed at higher levels in 11.5 dpc EGFP-positive male gonad cells compared with EGFP-positive female gonad cells (>2-fold, P<0.05; Fig. 4A). The 266 transcripts represented in both data sets are genes that are exclusively upregulated in 11.5 dpc male somatic gonad cells (Fig. 4A). These genes were sorted on fold change in 11.5 dpc male gonad samples compared with 10.5 dpc male gonad samples, and the first 40 genes are displayed in Table 2. Supplementary Material, Table S1 shows the entire list with male somatic gonad cells-specific genes. Supplementary Material, Table S2 shows the raw signal data of the Affymetrix screens.


Figure 4634
View larger version (21K):
[in this window]
[in a new window]
  		Figure 4. Identification of candidate genes by subtraction strategies. (A) Microarray screens identified 1873 genes (1607+266) upregulated in 11.5 dpc EGFP-positive male somatic gonad cells compared with 10.5 dpc EGFP-positive male somatic gonad cells, and 796 genes (530+266) upregulated in 11.5 dpc EGFP-positive male somatic gonad cells compared with 11.5 dpc EGFP-positive female somatic gonad cells. Two hundred and sixty-six genes were found to be upregulated exclusively in 11.5 dpc male somatic gonad cells. (B) 2426 genes (2184+242) were upregulated in 11.5 dpc EGFP-positive female somatic gonad cells when compared with 10.5 dpc EGFP-positive female somatic gonad cells and 512 genes (270+242) were upregulated in 11.5 dpc EGFP-positive female somatic gonad cells when compared with 11.5 dpc EGFP-positive male somatic gonad cells. Two hundred and forty-two genes were found to be upregulated exclusively in 11.5 dpc female somatic gonad cells.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Top 40 genes specifically upregulated in 11.5 dpc male somatic cells (>2-fold, P<0.05)
 
A number of genes identified in this experiment, including Dhh (38Go), Fgf9 (39Go) and Ptgds (2Go,40Go,41Go) (shown in bold in Table 2, Supplementary Material, Table S1, Fig. 5A–C) are known to have important functions during early male gonad development. These results indicate that the screen was successful in identifying genes important for sex determination. Sox9 was also identified, but was upregulated only 1.4-fold in 11.5 dpc male samples (data not shown) and, therefore, does not appear in Table 2 or Supplementary Material, Table S1. Curiously, Sry was not detected, even though we found clear evidence of Sry mRNA in 11.5 dpc male gonad RNA by quantitative PCR (Fig. 3). In contrast, steroidogenic markers Cyp11a1 and Hsd3b1 were not identified, indicating an absence of early differentiating Leydig cells in the green 11.5 dpc male somatic gonad cell sample.


Figure 4635
View larger version (17K):
[in this window]
[in a new window]
  		Figure 5. Expression profiles of known genes in 10.5 and 11.5 dpc male and female somatic gonad cell populations. (A) Fgf9, (B) Dhh, (C) Ptgds and (D) Fst. Sex (XX and XY) and age (10.5 and 11.5 dpc) are plotted on the X-axes. The Y-axes show raw intensity values on a linear arbitrary scale.

 
We next determined in silico the subcellular location of these 266 genes and found that many encode nuclear proteins including transcription factors, extracellular proteins including signalling factors and membrane associated proteins including membrane receptors. The largest proportion of genes has no known homology and their cellular role(s) cannot be predicted (Fig. 6A).


Figure 4636
View larger version (18K):
[in this window]
[in a new window]
  		Figure 6. Most of the up- or downregulated genes in 11.5 dpc male somatic gonad cells and upregulated genes in 11.5 dpc female somatic gonad cells are unclassified. (A) Cellular distribution of the 266 genes upregulated exclusively in 11.5 dpc male somatic cells (>2-fold, P<0.05). (B) Cellular distribution of the 50 genes downregulated exclusively in 11.5 dpc male somatic cells (>2-fold, P<0.05). (C) Cellular distribution of the 242 genes upregulated exclusively in 11.5 dpc female somatic cells (>2-fold, P<0.05). Description of main sections of the pie diagrams: ‘unknown’ includes genes with no annotation data, ‘cell surface’ includes genes encoding cell surface receptors, ‘extracellular’ includes genes encoding signalling molecules and extracellular matrix proteins and ‘nucleus’ includes genes encoding transcription factors.

 
Genes downregulated in male somatic cells only
Genetic evidence from analysis of human SRY-negative XX males has led to the suggestion that SRY may act as a transcriptional repressor (42Go). Genes identified in the present screen as being downregulated exclusively in 11.5 dpc male somatic gonad cells may, therefore, be genes that are repressed directly or indirectly by SRY. We found 43 genes with unchanged expression levels in 11.5 dpc female somatic gonad cells that were downregulated in 11.5 dpc male somatic gonad cells (>2-fold, P<0.05). In addition, seven genes were found to be upregulated in female somatic gonad cells (>2-fold, P<0.05) that were downregulated in 11.5 dpc male somatic gonad cells (>2-fold, P<0.05). These genes were pooled and the first 40 genes are displayed in Table 3. Supplementary Material, Tables S3 and S4 show the complete list of genes that are downregulated in 11.5 dpc male somatic gonad cells and the raw signal data of the Affymetrix screens. Figure 6B shows that most of these genes are unknown genes.


View this table:
[in this window]
[in a new window]
  		Table 3. Top 40 genes specifically downregulated in 11.5 dpc male somatic gonad cells (>2-fold, P<0.05)
 
Genes upregulated in female somatic gonad cells only
Genes that are exclusively upregulated in 11.5 dpc female somatic gonad cells (compared with female somatic gonad cells at 10.5 dpc and with male somatic gonad cells at 11.5 dpc) may include novel candidates for roles in the earliest steps of female gonad development and may include female sex-determining genes, if these indeed exist. In total, 2426 genes were found to be upregulated in EGFP-positive gonad cells of 11.5 dpc compared with 10.5 dpc female embryos (>2-fold, P<0.05; Fig. 4B) and 512 genes were expressed more highly in EGFP-positive gonad cells of 11.5 dpc female embryos compared with their male counterparts (>2-fold, P<0.05; Fig. 4B). The 242 transcripts represented in both data sets (Fig. 4B) are genes upregulated in somatic cells of 11.5 dpc female gonads only. These genes were sorted on fold change in 11.5 dpc female gonad samples compared with 10.5 dpc female gonad samples and the first 40 genes are displayed in Table 4. Supplementary Material, Tables S5 and S6 show the complete list of genes that are upregulated in 11.5 dpc female somatic gonad cells and the raw signal data of the Affymetrix screens. Among the first 20 listed genes is Fst, one of the few genes known to be upregulated in 11.5 dpc female gonads (shown in bold in Table 4 and Supplementary Material, Table S5, Fig. 5D; 18Go,43Go). Figure 6C shows that although many female gonad-specific gene products are nuclear, membrane-associated or extracellular proteins, more than half of the identified genes are novel with no known homology.


View this table:
[in this window]
[in a new window]
  		Table 4. Top 40 genes specifically upregulated in 11.5 dpc female somatic cells (>2-fold, P<0.05)
 
Validation of microarray data by quantitative PCR
To further validate the outcomes of our microarray screens, we performed quantitative PCR assays using RNA extracted from whole 11.5 dpc male and female genital ridges, and primer pairs listed in Supplementary Material, Table S7. We first assessed the expression levels of 11 genes shown by microarray analyses to be expressed with the highest fold differences in 11.5 dpc male somatic gonad cells (the top 11 genes in Table 2 and Supplementary Material, Table S1). Of these, six (Tesc, Cst9, Col9a3, Mmd2, Adamts16 and Dhh) showed significantly higher levels, two (Centb1 and Slc26a7) showed marginally higher levels and three (Afp, Ogn and Mapk13) showed similar levels of expression in male versus female gonads (Fig. 7 and data not shown). Because Afp is highly expressed in blood cells, which autofluoresce in FACS analyses, this gene appears to represent a false-positive outcome of our FACS-based screening method, caused by contamination of sorted cell samples with blood cells. In other cases, disparity between expression fold differences in sorted cells (Table 2) and differences measured by quantitative PCR are likely to be due to the use of whole gonads, containing multiple cell types, in the PCR analyses. Overall, the expression analyses indicate that the majority of genes identified in our screen as being male-specific are valid candidates for further analysis and that the proportion of false-positive genes is low.


Figure 4637
View larger version (33K):
[in this window]
[in a new window]
  		Figure 7. Validation of microarray experiments by quantitative PCR assays. The top row shows the expression levels in 11.5 dpc male (dark bars) and female (light bars) genital ridges of 10 genes that in microarray analyses were found to be expressed to the highest fold differences in somatic cells of male gonads compared with somatic cells of female gonads. The bottom row shows the expression levels in 11.5 dpc male (dark bars) and female (light bars) genital ridges of seven genes that in microarray analyses were found to be expressed to the highest fold differences in somatic cells of female gonads compared with somatic cells of male gonads. The Y-axes display relative fold expression levels compared with 18S rRNA. Experiments were carried out at least three times and each experiment consisted of three replicates. Representative experiments are shown in the figure. The error bars represent minimal and maximal ranges of gene expression levels in three replicates when compared with 18S rRNA.

 
We next tested by quantitative PCR the expression levels of eight genes that were expressed with the highest fold differences in 11.5 dpc female somatic gonad cells (the top eight in Table 4 and Supplementary Material, Table S5). Of these, six (Klk6, 6430710C18Rik, Sprr2d, AW124722, D12Ertd647e and Ugt2b5) were expressed at significantly higher levels in whole 11.5 dpc female gonads compared with male gonads (Fig. 7). Only one (2210407C18Rik) showed no reproducible difference in expression levels between male and female genital ridges. These data further support the validity of the microarray screens.

Identification of human orthologues and chromosomal loci
Next, we set out to identify the human orthologues of the mouse genes that were found in our microarray screens and queried the BioMart data mining tool (www.ensembl.org). Results are displayed in Tables 2–4. Interestingly, we detected many human orthologues that map to chromosomal loci associated with human sex disorders ranging from severe intersex and sex-reversal syndromes to relatively mild dysgenesis syndromes such as hypospadias and cryptorchidism (44Go).

Notably, ADAM8, PIK3AP1, PNLIPRP1, SFXN3, SLC18A2, TRIM47 and some novel genes (Tables 2–4, Supplementary Material, Tables S1, S3 and S5) map to human chromosome 10q, a region associated with complete sex reversal and ambiguous genitalia (45Go). KRTAP17-1, RND2, THRAP1 and a novel gene (Tables 2–4, Supplementary Material, Tables S1, S3 and S5) map to the region 17q21–q24, associated with Meckel syndrome type I, characterized by genital abnormalities ranging from hypospadias to sexual ambiguity (46Go–50Go). Finally, several genes identified in this study map to loci that have been associated with cryptorchidism and hypospadias, such as Xq12–q21.31 (FG syndrome) and 16p13 (Rubinstein–Taybe syndrome) (51Go). However, it should be kept in mind that mild sex disorders such as hypospadias and cryptorchidism may have genetic causes, unrelated to malfunction of gonad development.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
In this study, we generated and characterized a novel transgenic reporter mouse line in which EGFP expression was driven by a 674 bp Sf1 promoter fragment, enabling us to carry out expression-profiling assays on flow-sorted cell populations during the critical interval of sex determination between 10.5 and 11.5 dpc. We showed that the cells expressing EGFP during this early period are somatic cells strongly enriched for cells of the supporting cell lineage. These cells are known to be the site of SRY activity and hence critical for sex determination and subsequent testis organization in the male, and assumed to have an analogously important role in the female. Using this system, we identified a large number of novel genes that can be considered strong candidates for a prominent role in sex determination and differentiation and associated disorders in humans.

SF1 is an important regulator of endocrine development. It controls the expression of Cytochrome p450 hydroxylases, Amh (or Mis) and Dax1 (52Go,53Go). Sf1 is expressed in tissues at multiple levels of the reproductive axis including the adrenal cortex, the gonads and the diencephalon (54Go). Within the urogenital ridge, Sf1 transcripts are first detected at 9.0 dpc in a continuous population of cells but it is not until 11.0 dpc that discrete and separate populations of gonadal and adrenal precursor cells are first distinguishable (54Go). In other studies, in a transgenic reporter mouse line in which expression of EGFP was placed under control of a 50 kb genomic fragment of the mouse Sf1 gene, the EGFP expression pattern was found to closely follow the endogenous Sf1 expression pattern in most tissues, with the striking exception of the pituitary primordium, where EGFP expression was not detected (55Go). In our studies, however, we found that EGFP was expressed exclusively in gonadal cells and not in the adrenal primordium or in the diencephalon. This discrepancy in expression patterns is unlikely to be dependent on random effects of the transgene-integration site, as we and others have observed this absence of EGFP expression in the adrenal or diencephalon in all transgenic mouse lines generated with this promoter fragment (33Go). Therefore, we can conclude that the WT1- and LHX9-binding sites present within the 674 bp promoter fragment of Sf1 are most likely not the enhancers required for directing expression of Sf1 into adrenal cells, which must be under the control of other elements, located outside of this promoter fragment. Importantly, this specificity of expression also assures that the EGFP-positive cell populations that were used in our studies are not contaminated with precursor cells of the adrenal anlage.

A number of lines of evidence suggest that Sf1 marks only cells of the supporting cell lineage (pre-Sertoli and pre-granulosa cells) at the stages examined in this study. Our analysis of transgene expression in 10.5 and 11.5 dpc embryos reveals an abundance of positive cells in the genital ridges at this early stage, whereas Leydig and other stromal cells are known to immigrate from the mesonephros after this time point (12Go,14Go). However, in the absence of the mesonephros, Leydig cells can still differentiate within XY gonads, indicating that the mesonephros is not the only source for Leydig cells (11Go). Schmahl et al. (9Go) found a second source of Leydig cell precursors within the coelomic epithelium of XY gonads, but were able to determine that SF1-expressing coelomic epithelial cells of the gonad labelled before 11.5 dpc exclusively differentiated into Sertoli cells, whereas a second population of proliferating but SF1-negative coelomic epithelial gonad cells labelled during a slightly later phase gave rise only to Leydig cells. Therefore, Leydig cell precursors coming from the coelomic epithelium do not express SF1 until after 11.5 dpc.

These findings suggest that EGFP-positive gonad cell populations of both 10.5 and 11.5 dpc embryos of our pSf1–EGFP transgenic reporter line consist of supporting cells (pre-Sertoli and pre-granulosa cells) and not Leydig cell precursors. In support of this, we did not detect the earliest markers of Leydig cell differentiation, Cyp11a1 and 3ßHsd, among the 266 identified 11.5 dpc male somatic gonad cell-specific genes. We also found that pre-Sertoli cells within the sex cords of 12.5 dpc transgenic testes displayed higher levels of EGFP expression than Leydig cells located in the gonad interstitium. Together, these observations suggest that the FACS-sorted EGFP-positive cell populations were strongly enriched for pre-Sertoli cells and, therefore, that the expression profiles of 10.5 and 11.5 dpc male gonad EGFP-positive cells reflect the transcriptional changes that occur within the supporting cell lineage, the critical lineage for sex determination. Therefore, all male genes identified in our screen can be considered as potential (direct or indirect) downstream targets of SRY and/or activators of Sox9 expression. It remains to be investigated by molecular studies whether this is indeed the case.

Importantly, our screens identified several genes with known functions in early male and female gonad development, such as Dhh, Fgf9, Ptgds and Fst (shown in bold in Tables 2 and 4, Fig. 5). Dhh has been shown in gene-targeting studies in mice to be important for differentiation of both Sertoli and Leydig cells (38Go), Fgf9 is thought to act downstream of Sry, but upstream of Sox9 during the process of sex determination, and Fgf9 loss-of-function causes sex reversal in mice (39Go,56Go). Ptgds, responsible for synthesis of prostaglandin D2, recently has been shown to be produced by pre-Sertoli cells immediately after Sry to recruit more gonad cells into the Sertoli cell lineage (2Go,40Go,41Go). Fst is expressed in somatic cells of the female gonad and was shown to act downstream of WNT4 during ovary development (18Go,43Go). This indicates that our screens were effectively able to identify genes with important described roles during the earliest steps of sex determination and male and female gonad development.

Moreover, our microarray screens also identified a number of genes with yet unknown roles that have previously been detected in other expression screens as being expressed to higher levels in—generally older—male or female gonads (25Go–32Go). The expression patterns of a number of genes that show the most striking changes in expression levels between samples, namely Aard, Cbln4, Col9a2, Col9a3, Cst9, Dtna, Mmd2, Serpine2 and Tesc (shown in bold in Table 2) have been studied by others in more detail. Thus, it has been demonstrated that these genes are indeed expressed in pre-Sertoli cells of male gonads (27Go,57Go–60Go), thereby further attesting to the validity of our methods.

One gene that was not detected by our gene expression-profiling studies in the 11.5 dpc EGFP -positive male gonad cell population is Sry, even though we were able to detect Sry messenger RNA in these cells using quantitative PCR (Fig. 3). There are a number of possible explanations for this observation. First, in the quantitative PCR experiments, random primers were used for the generation of cDNA, which may have been more efficient than the oligo(dT) primers that were used during the first round of RNA amplification. Second, signals detected by different probe sets present on the Affymetrix genome array for the same gene often do not yield the same signals, indicating that not all probe sets are equally efficient in binding aRNA. It is possible that the two Sry probe sets (1450578_at and 1450579_x_at) present on the Mouse Genome 430 2.0 arrays are not very efficient in binding to the aRNA encoding Sry.

The underlying genetic cause of many human sex disorders is not understood. In many cases, an affected chromosomal locus has been described, but the genes responsible for the malformations have not been identified. Overall, in 75% of the human cases of sex reversal, the genetic cause remains unknown (24Go). We found that many human orthologues of the identified mouse genes map to genomic regions that have been associated with human sexual dysgenesis syndromes, making them excellent candidate disease genes (44Go). In addition, given the nature of our microarray screens, it seems likely that there may be other important candidate intersex genes among the human gene orthologues listed in Tables 2–4. Further investigation by functional analyses in mice and mutation searching in human patient DNA will be required to test this prediction.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Mouse handling
Protocols and use of animals in this experiment were approved by the Animal Welfare Unit of the University of Queensland, registered as an institution that uses animals for scientific purposes under the Queensland Animal Care and Protection Act (2001). Timed matings were carried out and 0.5 dpc was defined as noon on the day of discovery of the copulation plug. Mouse embryos were staged according to limb bud morphology (61Go).

Generation of the pSf1–EGFP transgenic mouse line
A 674 bp fragment of the Sf1 promoter (33Go) and cDNA-encoding EGFP were cloned by PCR into pTransgene, a vector designed for efficient generation of highly expressing transgenic mice (kind gift of Dr David Sherr, Boston, MA, USA; Accession no. AF515846 [GenBank] ; for cloning primer sequences, see Supplementary Material, Table S7). Both inserts of the transgenic vector were sequenced. The transgene was excised with restriction enzymes PacI, EcoRI and BglI, and transgenic founders were generated by standard pronuclear injection into zygotes of F2 offspring of CBAxC57BL6 crosses. Founders were genotyped by PCR using primers that amplify EGFP cDNA (Supplementary Material, Table S7).

Embryo imaging and immunofluorescence
Embryos were dissected and studied under a dissection microscope containing a mercury lamp exciting EGFP fluorescence and photographed under bright light exposure and subsequently under UV exposure. Files of UV light exposures were inverted and pasted into the green channels of RGB bright-field exposures using Photoshop 7.0 (Adobe Systems, Inc.). Embryos were fixed in 4% paraformaldehyde at 4°C overnight, washed with phosphate-buffered saline (PBS), incubated overnight in 30% sucrose in PBS at 4°C and then embedded in OCT compound (Tissue-Tek). Embedded embryos were cut in serial sections of 12 µm thickness using a Leica cryostat. Cryosections were washed in PBS+0.1% Tween-20 (PBST), pre-blocked in 10% heat-inactivated sheep serum (HISS) in PBST and incubated overnight at 4°C with primary antibody solutions in 10% HISS in PBST in a humidified chamber. Rabbit anti-SF1 antibody was used at 1:200 and was a kind gift of Dr Ken-Ichirou Morohashi (Okazaki, NIBB, Japan). Rat anti-PECAM1 was used at 1:200 (purchased from BM Pharmingen). Rabbit anti-laminin was used at 1:200 (Sigma). Subsequently, samples were washed in PBST and blocked in 10% HISS in PBST, followed by incubation in a humidified chamber with 1:200 dilution of secondary antibodies (anti-rat Alexa594 and anti-rabbit Alexa594; Molecular Probes) for 2 h at room temperature. Sections were mounted and examined using a Zeiss LSM 510 Meta confocal microscope.

Cell dissociation and FACS
Gonads at 10.5 and 11.5 dpc were dissected in Leibovitz's L15 medium (Gibco) and sexed by PCR using Zfy primers (Supplementary Material, Table S7). Male and female gonads of 10.5 and 11.5 dpc, respectively, were pooled in four separate tubes. Gonad cells were enzymatically dissociated in Hank's balanced salt solution (Sigma) containing 1 mg/ml collagenase B (Roche), 1.2 U/ml Dispase II (Roche) and 5 U/ml DNase1 (Sigma) for 20 min at 37°C while shaking. Cells were further dissociated mechanically using a P1000 Gilson pipette and a 23-gauge syringe in dissociation medium, with two intermittent 5 min incubation steps at 37°C. Finally, cells were passed through 40 µm cell strainers (Falcon BD Biosciences), rinsed with ice-cold PBS, spun down and resuspended using a 23-gauge syringe in 2 ml of ice-cold PBS. All cell-sorting experiments were carried out at the Queensland Brain Institute of the University of Queensland, Brisbane, Australia, using a BD FACS Vantage SE with FACSDiVa Option, which was run at 28 psi using a 90 µm nozzle. EGFP fluorochromes were excited with a 488 nm argon laser and 530 nm collection filters were used to detect EGFP. GFP-positive and GFP-negative gonad cell populations were collected in separate tubes and kept on ice before further processing.

RNA extraction, primer design and real-time PCR
Total RNA was extracted from FACS-sorted cell populations, whole 11.5 dpc genital ridges and 13.5 dpc gonads using RNeasy Micro kit (Qiagen) according to the manufacturer's instructions and including the optional DNase I genomic DNA degradation step. Total RNA was reverse-transcribed using random hexamer primers (Promega) and Superscript III (Invitrogen), using standard methodologies.

Real-time PCR primer pairs were designed with Tm of close to 60°C to generate 60–145 bp amplicons mostly spanning introns or using the Universal Probe library tool on the Roche Web site (http://www.roche-applied-science.com/sis/rtpcr/upl/adc.jsp) (Supplementary Material, Table S7 and J. Bowles and P.K., unpublished data). All primer pairs were validated to demonstrate that efficiencies of target cDNA and reference 18S amplification were approximately equal.

Quantitative PCR assays were carried out using SYBR green PCR master mix and validated primer pairs using an ABI PRISM® 7000 Sequence Detection System following the manufacturer's instructions. Target cDNA levels were analysed by the comparative cycle time (Ct) method of real-time RT–PCR and values were normalized to 18S rRNA expression levels.

RNA amplification
RNA was linearly amplified in two consecutive rounds using a MessageAmpTMII aRNA kit (Ambion) following the manufacturer's instructions. Second round aRNA was labelled with biotin-11-CTP and biotin-16-UTP, quantified using a spectrophotometer. aRNA quality was assessed by interrogating test3 microarray chips (Affymetrix) according to the manufacturer's instructions to confirm that quality and bias for all samples were comparable. aRNA was fragmented in fragmentation buffer for 5 min at 95°C and stored at –80°C until further use.

Microarray processing and analysis
All microarray processing was carried out at the Queensland Brain Institute, the University of Queensland, using mouse genome 430 2.0 arrays (Affymetrix). These high-density oligonucleotide arrays comprise over 45 101 probe sets representing over 34 000 well-substantiated mouse genes (http://www.affymetrix.com/support/technical/libraryfilesmain.affx). Fragmented aRNA (10–15 µg) was hybridized to microarray slides (‘chips’) in triplicate and hybridized for 16 h at 45°C. Chips were washed and stained with streptavidin phycoerythrin using fluidics protocol EukGE-WS2v5 and a GeneChip Fluidics Station 450 (Affymetrix) according to the manufacturer's standard protocols. The arrays were scanned using a GeneChip scanner 3000 (Affymetrix). The output was analysed using GeneChip Operating Software v1.2 (GCOSv1.2, Affymetrix) and examined for excessive background and evidence of RNA degradation. All experiments were scaled to a target intensity of 150.

Scanned images of the arrays were converted by GCOSv1.2 and imported into GeneSpring 7.2 (Silicon Genetics, Redwood City, CA, USA) for further analysis. Data were normalized per chip and per probe set using default normalization methods, including setting of signal values≤0.01, total chip normalization to the 50th percentile and normalization of each gene to the median with cut off value 0.01. Differential expression was defined as a difference of 2-fold or greater in all replicates when comparing two samples. Statistical analyses were performed using GeneSpring software by using the Student's t-test (significance level set at P<0.05).

Database searches
Mouse 430 2.0 Affymetrix Genome Array IDs were used to query the NetAffx data mining tool (www.affymetrix.com/analysis/index.affx) for gene annotations. The BioMart data mining tool (www.ensembl.org) was queried for human orthologues, Entrez Gene IDs, human chromosomal loci and MIM IDs.


   NOTE ADDED IN PROOF
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 NOTE ADDED IN PROOF
SUPPLEMENTARY MATERIAL
 REFERENCES
 
During the review of this manuscript, an article was published describing microarray profiling of mouse fetal somatic gonadal cells using an Sf1–EGFP reporter mouse line (62Go). The outcomes of that study are comparable to the data reported here.


   SUPPLEMENTARY MATERIAL
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
REFERENCES
 
Supplementary Material is available at HMG Online.


   ACKNOWLEDGEMENTS
 
The authors thank Jo Bowles, Alex Combes, Andrew Jackson and Grant Challen for critically reading the manuscript and Andrew Jackson, Grant Challen, Kelly Hanson, Jo Bowles, Sean Grimmond, Kyra Woods and Virginia Nink for their invaluable help during the course of the study. We are grateful to Ken-Ichirou Morohashi, Toshiaki Noce and David Sherr for provision of reagents. We thank Tara Davidson, Elizabeth Williams, Laurence Venness, Linda Wernbacher, Joanne Dowd, Anne Hardacre, David McNeilly and Michelle Kappler for generation and maintenance of the transgenic stocks and members of group Koopman for help with dissections and stimulating discussions. We also thank the anonymous referees for their helpful suggestions to improve the manuscript. This work was supported by grants from the Australian Research Council (ARC) and the National Health and Medical Research Council of Australia. P.K. is an ARC Australian Professorial Research Fellow.

Conflict of Interest statement. None declared.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 NOTE ADDED IN PROOF
 SUPPLEMENTARY MATERIAL
 REFERENCES
 

  1. Sekido, R., Bar, I., Narvaez, V., Penny, G. and Lovell-Badge, R. (2004) SOX9 is up-regulated by the transient expression of SRY specifically in Sertoli cell precursors. Dev. Biol., 274, 271–279.[CrossRef][ISI][Medline]

  2. Wilhelm, D., Martinson, F., Bradford, S., Wilson, M.J., Combes, A.N., Beverdam, A., Bowles, J., Mizusaki, H. and Koopman, P. (2005) Sertoli cell differentiation is induced both cell-autonomously and through prostaglandin signaling during mammalian sex determination. Dev Biol., 287, 111–124.[CrossRef][ISI][Medline]

  3. Albrecht, K.H. and Eicher, E.M. (2001) Evidence that Sry is expressed in pre-Sertoli cells and Sertoli and granulosa cells have a common precursor. Dev. Biol., 240, 92–107.[CrossRef][ISI][Medline]

  4. Hacker, A., Capel, B., Goodfellow, P. and Lovell-Badge, R. (1995) Expression of Sry, the mouse sex determining gene. Development, 121, 1603–1614.[Abstract]

  5. Bullejos, M. and Koopman, P. (2001) Spatially dynamic expression of Sry in mouse genital ridges. Dev. Dyn., 221, 201–205.[CrossRef][ISI][Medline]

  6. Jeske, Y.W., Bowles, J., Greenfield, A. and Koopman, P. (1995) Expression of a linear Sry transcript in the mouse genital ridge. Nat. Genet., 10, 480–482.[CrossRef][ISI][Medline]

  7. Kent, J., Wheatley, S.C., Andrews, J.E., Sinclair, A.H. and Koopman, P. (1996) A male-specific role for SOX9 in vertebrate sex determination. Development, 122, 2813–2822.[Abstract]

  8. Morais da Silva, S., Hacker, A., Harley, V., Goodfellow, P., Swain, A. and Lovell-Badge, R. (1996) Sox9 expression during gonadal development implies a conserved role for the gene in testis differentiation in mammals and birds. Nat. Genet., 14, 62–68.[CrossRef][ISI][Medline]

  9. Schmahl, J., Eicher, E.M., Washburn, L.L. and Capel, B. (2000) Sry induces cell proliferation in the mouse gonad. Development, 127, 65–73.[Abstract]

  10. Karl, J. and Capel, B. (1998) Sertoli cells of the mouse testis originate from the coelomic epithelium. Dev. Biol., 203, 323–333.[CrossRef][ISI][Medline]

  11. Merchant-Larios, H. and Moreno-Mendoza, N. (1998) Mesonephric stromal cells differentiate into Leydig cells in the mouse fetal testis. Exp. Cell Res., 244, 230–238.[CrossRef][ISI][Medline]

  12. Nishino, K., Yamanouchi, K., Naito, K. and Tojo, H. (2001) Characterization of mesonephric cells that migrate into the XY gonad during testis differentiation. Exp. Cell Res., 267, 225–232.[CrossRef][ISI][Medline]

  13. Buehr, M., Gu, S. and McLaren, A. (1993) Mesonephric contribution to testis differentiation in the fetal mouse. Development, 117, 273–281.[Abstract]

  14. Martineau, J., Nordqvist, K., Tilmann, C., Lovell-Badge, R. and Capel, B. (1997) Male-specific cell migration into the developing gonad. Curr. Biol., 7, 958–968.[CrossRef][ISI][Medline]

  15. Merchant-Larios, H., Moreno-Mendoza, N. and Buehr, M. (1993) The role of the mesonephros in cell differentiation and morphogenesis of the mouse fetal testis. Int. J. Dev. Biol., 37, 407–415.[ISI][Medline]

  16. Swain, A., Zanaria, E., Hacker, A., Lovell-Badge, R. and Camerino, G. (1996) Mouse Dax1 expression is consistent with a role in sex determination as well as in adrenal and hypothalamus function. Nat. Genet., 12, 404–409.[CrossRef][ISI][Medline]

  17. Vainio, S., Heikkila, M., Kispert, A., Chin, N. and McMahon, A.P. (1999) Female development in mammals is regulated by Wnt-4 signalling. Nature, 397, 405–409.[CrossRef][Medline]

  18. Yao, H.H., Matzuk, M.M., Jorgez, C.J., Menke, D.B., Page, D.C., Swain, A. and Capel, B. (2004) Follistatin operates downstream of Wnt4 in mammalian ovary organogenesis. Dev. Dyn., 230, 210–215.[CrossRef][ISI][Medline]

  19. Loffler, K.A., Zarkower, D. and Koopman, P. (2003) Etiology of ovarian failure in blepharophimosis ptosis epicanthus inversus syndrome: FOXL2 is a conserved, early-acting gene in vertebrate ovarian development. Endocrinology, 144, 3237–3243.[Abstract/Free Full Text]

  20. Loffler, K.A. and Koopman, P. (2002) Charting the course of ovarian development in vertebrates. Int. J. Dev. Biol., 46, 503–510.[ISI][Medline]

  21. Byskov, A.G. (1978) The anatomy and ultrastructure of the rete system in the fetal mouse ovary. Biol. Reprod., 19, 720–735.[Abstract]

  22. Odor (1969) Ultrastructural studies on fetal and early post-natal mouse ovaries. 1. Histogenesis and organogenesis. Am. J. Anat., 124, 163–186.[CrossRef][ISI][Medline]

  23. Pepling, M.E. and Spradling, A.C. (2001) Mouse ovarian germ cell cysts undergo programmed breakdown to form primordial follicles. Dev. Biol., 234, 339–351.[CrossRef][ISI][Medline]

  24. Vilain, E. and McCabe, E.R. (1998) Mammalian sex determination: from gonads to brain. Mol. Genet. Metab., 65, 74–84.[CrossRef][ISI][Medline]

  25. Wertz, K. and Herrmann, B.G. (2000) Large-scale screen for genes involved in gonad development. Mech. Dev., 98, 51–70.[CrossRef][ISI][Medline]

  26. Bowles, J., Bullejos, M. and Koopman, P. (2000) A subtractive gene expression screen suggests a role for vanin-1 in testis development in mice. Genesis, 27, 124–135.[CrossRef][ISI][Medline]

  27. Menke, D.B. and Page, D.C. (2002) Sexually dimorphic gene expression in the developing mouse gonad. Gene Expr. Patterns, 2, 359–367.[CrossRef]