This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Roy, A. Articles by Matzuk, M. M Search for Related Content PubMed PubMed Citation Articles by Roy, A. Articles by Matzuk, M. M

Deconstructing mammalian reproduction: using knockouts to define fertility pathways

¹ Departments of Pathology, ² Molecular and Human Genetics and ³ Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA

Correspondence should be addressed to Martin M Matzuk at the Department of Pathology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA; Email: mmatzuk{at}bcm.tmc.edu

	Abstract

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Reproduction is the sine qua non for the propagation of species and continuation of life. It is a complex biological process that is regulated by multiple factors during the reproductive life of an organism. Over the past decade, the molecular mechanisms regulating reproduction in mammals have been rapidly unraveled by the study of a vast number of mouse gene knockouts with impaired fertility. The use of reverse genetics to generate null mutants in mice through targeted disruption of specific genes has enabled researchers to identify essential regulators of spermatogenesis and oogenesis in vivo and model human disorders affecting reproduction. This review focuses on the merits, utility, and the variations of the knockout technology in studies of reproduction in mammals.

	Introduction

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Living forms are defined by their ability to reproduce. Although different organisms produce progeny using diverse methods (asexual and sexual reproduction, parthenogenesis, etc.), the basic principles of reproduction, both at the molecular and the organismal levels, have remained the same since the dawn of life: formation of gametes to facilitate the transfer of genetic material and epigenetic information to the next generation. Indeed, in several model organisms, orthologous genes have been identified that perform similar roles in reproductive physiology, thereby highlighting a conservation of this complex biological process.

Historically, two major obstacles have impeded the study of reproductive biology in mammals (reviewed by Staub 2001). First, germ cells in mammals develop in a microenvironment of supporting stromal cells of somatic origin that interacts with the former through autocrine/paracrine mechanisms as well as direct cell-to-cell interactions (Matzuk et al. 2002). Devoid of this support, isolated germ cells in culture fail to survive and maintain their characteristics, making it highly difficult to study molecular interactions and pathways that are unique to these cells. The second, but related, hindrance has been the inability to develop a proper cellular context in vitro that would allow undifferentiated germ cells to differentiate into specialized gametes that retain their ability to fertilize and produce healthy offspring. Recently, interest has exploded in the field of in vitro germ cell differentiation (Lacham-Kaplan 2004, West & Daley 2004) due to exciting findings by several investigators who have been able to generate meiosis-competent germ cells from undifferentiated embryonic stem cells (Hubner et al. 2003, Toyooka et al. 2003, Geijsen et al. 2004). Although these gametes have been shown to fertilize oocytes and produce blastocysts, no data are available yet to suggest that these can progress to produce live offspring. In addition, Feng and coworkers (2002) have generated an undifferentiated spermatogonial cell line immortalized by the ectopic expression of the catalytic component of telomerase. The cell line was also shown to differentiate in vitro to morphologically similar haploid round spermatids but the authors did not investigate whether these haploid germ cells were competent to initiate fertilization and zygotic development when injected into oocytes. So, although the future looks promising, more careful analysis of the physiological state of these in vitro-formed gametes, including the identification of parent-of-origin-specific imprints, would be critical for enhancing the utility of these techniques.

Our understanding of the molecular mechanisms of mammalian reproduction therefore has been largely dependent on loss-of-function mutagenesis in mice. Ever since researchers developed techniques to culture mouse embryonic cells in vitro (Evans & Kaufman 1981) and methods were developed to inject cultured embryonic pluripotential cells into blastocysts to produce germ-line chimeras (Bradley et al. 1984), the field of mouse transgenic technology has undergone a veritable explosion. Around the same time, researchers were successfully targeting exogenous DNA into endogenous loci based on limited amounts of homology between the two DNA sequences in cultured mammalian cells (Smithies et al. 1985). Simultaneously, Thomas and colleagues (Thomas et al. 1986, Thomas & Capecchi 1987) had developed a highly efficient method of targeted mutagenesis of cultured embryonic stem cells and other mammalian cells by using homologous recombination between endogenous genes and exogenously introduced DNA (targeting vector), a technique that was shown to be highly specific by Doetschman and coworkers (Doetschman et al. 1987, 1988). Our present-day knockout technology, a cornerstone of functional genomic studies, got its final form when these two technologies were harnessed together to generate the first targeted knockout mouse model through homologous recombination in embryonic stem cells (Koller et al. 1989).

A knockout mouse model is typically designed to produce a null mutation in the target gene and thus uncover an essential function. In dissecting gene functions, a hypomorphic mutation or the generation of an allelic series can at times be beneficial to researchers testing for genetic interactions between candidate genes. However, a null mutation is preferred in identifying non-redundant and essential functions of genes, as exemplified by the generation of approximately 300 mouse knockouts that manifest defects in reproduction (see below and online table that accompanies Matzuk & Lamb 2002).

Among the earliest mouse knockouts generated that were found to affect different aspects of reproduction were Inha (inhibin ), Lif (leukemia inhibitory factor), and estrogen receptor (Matzuk et al. 1992, Stewart et al. 1992, Lubahn et al. 1993). While LIF-deficient females had defects in implantation, knockout males and females of inhibin were infertile and had gonadal tumors. Female mice null for estrogen receptor were infertile, and males had reduced fertility. Over the last decade, several hundred mouse knockouts have been created that affect different stages of reproduction (reviewed in Matzuk & Lamb 2002), including primordial germ cell (PGC) formation and migration into the gonads, gonad development and sex determination, postnatal gametogenesis, sperm–egg binding and fertilization, and early embryonic development. For more comprehensive background on these topics and the mouse models that have shed light on these stages of reproductive development, readers are referred to these excellent reviews: Bendel-Stenzel et al.(1998), Matzuk & Lamb (2002), Matzuk et al.(2002), Primakoff & Myles (2002), Zhao & Garbers (2002), de Rooij & de Boer (2003), Brennan & Capel (2004) and Park & Jameson (2005).

In this review, we focus, in a stage-specific manner, on specific knockouts that have proved pivotal in our understanding of the biology of reproduction, with special emphasis on a few in vivo models that have clarified or refuted conflicting findings from in vitro experiments. We then briefly describe the principle of conditional mutagenesis and highlight its use in reproductive studies. Finally we discuss some of the current deficits of these technologies that have tempered the enthusiasm of researchers embracing them as their primary tools in functional genomics.

	Germ-line specification: what seals the fate?

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
In most model organisms studied so far, the germ line is specified in embryos by a maternally derived source of egg cytoplasm referred to as the germ plasm. In mammals, however, germ-cell determination appears to be regulated within the classical developmental paradigm of cell signaling and cell adhesion.

In mice, PGCs are the founder cells of the germ line that are allocated from a pool of proximal epiblast cells at around embryonic day 7.2 (E 7.2; reviewed in Zhao & Garbers 2002). By using mice that were deficient for bone morphogenetic protein (BMP), member 4 (Bmp4 ^–/– mice), Lawson and colleagues (1999) provided clear evidence that germ-cell competence is induced in a subset of proximal epiblast cells in response to BMP4 secreted from the extra-embryonic ectoderm. In Bmp4 ^–/– mice PGCs were completely absent, and in Bmp4 heterozygotes the PGC number was either greatly reduced or completely absent, suggesting that the effect of BMP4 is dose-dependent (Lawson et al. 1999). Deficiency of the related transforming growth factor ß (TGFß ) superfamily members, BMP8b and BMP2, led to a very similar phenotype of reduced or absent PGCs on a mixed genetic background (Ying et al. 2000; Ying & Zhao 2001). Interestingly, mice null for SMAD5 or SMAD1 (Chang & Matzuk 2001, Tremblay et al. 2001), the intracellular effectors of BMP signaling, have pronounced defects in PGC formation as well. More recently, knockout mice carrying a targeted mutation in the type I BMP receptor, activin receptor-like kinase 2 (ALK2), were found to have a complete absence of PGCs, further highlighting the importance of this signaling pathway in germ-cell specification (de Sousa Lopes et al. 2004).

While these studies suggest that an intact BMP signaling system in the embryonic stages is necessary for PGC fate specification, what is not clear is whether such a signaling mechanism is actually sufficient to induce PGC fate. What are the primary targets of BMP4 and BMP8b signaling in the epiblast? Do they signal directly to the PGC precursors? Is there a master regulator of PGC fate? These are some of the questions that remain unsolved. However, a recent study, utilizing a knockout in the mouse zinc-finger-containing transcriptional repressor PRDM1 (PR domain containing 1, with zinc-finger domain), has uncovered an essential role for this protein in PGC specification. Embryos lacking the PRDM1 protein have a complete absence of PGCs while heterozygotes have a drastic reduction in PGC number (Vincent et al. 2005). Since Prdm1 transcripts are expressed in PGC precursors as early as E 7.0, this is the first transcriptional factor to be identified that plays an essential role in PGC specification. So, is PRDM1 the master regulator that seals the fate of PGC precursors into entering a germ-cell lineage? The answer to that question awaits further studies, which would address key questions regarding the sufficiency of PRDM1 in specifying a germ-cell fate upon non-committed epiblast cells and whether its action is cell-autonomous in these precursors.

Studying gene function through knockouts has proved highly significant in the field of reproduction not only because of the uncovering of new genes and pathways but also because these in vivo studies on many occasions served to prove or refute several models that were based on either cell-culture-based assays or on evidence from forward-genetics approaches. One instance is noteworthy: by analyzing the expression pattern of genes from solitary nascent PGCs, Saitou and colleagues (2002) proposed a mechanism for acquisition of germ-cell competence and specification in PGC precursors in the epiblast which was based on essential roles for two highly expressed and unique cDNAs (named fragilis and stella) in PGC fate determination. However, when the stella gene was knocked out in mice, there appeared to be no defects in the specification, maintenance or proliferation of PGCs (Payer et al. 2003, Bortvin et al. 2004); instead, knockout males were fertile although null females failed to sire normal numbers of offspring as the gene fulfilled an essential role in the pre-implantation development of embryos as a maternally derived factor. In this context it would be interesting to know the phenotype of mice with a null mutation in fragilis.

	PGC proliferation, migration and survival

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Subsequent to their specification from precursor cells, PGCs undergo several rounds of proliferation till E 13.5, around when they enter cell-cycle arrest in G₀ within the male gonad or meiotic prophase I in the female gonad. In between, they migrate to the gonads by E 10.5 (Molyneaux & Wylie 2004) after traveling through the hindgut and the dorsal mesentery in response to secreted cues from the environment.

Through knockout approaches, researchers have identified several new genes that play essential roles in some or all of these processes. Targeted disruption of a peptidylprolyl isomerase-encoding gene, Pin1, uncovered an unexpected role for this gene in the regulation of the cell cycle and PGC proliferation. On an isogenic C57BL/6J background, mice null for Pin1 were found to have reduced PGC numbers during the mid-to-late migratory phases (E 9.5–12.5) compared with wild-type mice (Atchison et al. 2003); by analyzing BrdU labeling, this reduction seemed to be due to a delay in progression through the cell cycle rather than due to a cell-cycle arrest or enhanced apoptosis in these cells.

Specific genes regulate the survival of the migratory germ cells: the POU domain transcription factor, Oct4 (Pou5f1; Mouse Genome Informatics), is a critical determinant of pluripotency in the inner cell mass that causes embryonic death when mutated (Nichols et al. 1998, Niwa et al. 2000). To circumvent the early embryonic death and to examine the effect of deletion of Oct4 from PGCs, investigators have used TNAP ^Cre (tissue non-specific alkaline phosphatase – Cre), which encodes the Cre recombinase only in the PGC population prior to E 10.5 (Kehler et al. 2004; Table 1). Surprisingly, loss of OCT4 protein from PGCs leads not to differentiation but to enhanced apoptosis in these cells prior to their colonization of the gonads.

PGCs that successfully navigate to the genital ridges enter cell-cycle arrest, although at different stages in the two sexes, as mentioned above. Heterozygosity for the transcription factor zinc-finger protein 148 (Zfp148) has been shown to increase apoptosis and cause complete loss of germ cells from the male gonad at the E 13.5 stage (Takeuchi et al. 2003), an effect that is probably due to improper regulation of the p53 protein in these cells. Thus Zfp148 is one more addition to the growing list of genes that seem to affect PGC specification and maintenance in a dosage-dependent manner in mice.

	Sex determination

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Acquisition of reproductive competence in mammals involves the development of both germ cells and the supporting somatic cells that together constitute the gonad. The development of either the testis or the ovary from a bipotential gonad is a complex developmental process that is influenced by several transcriptional pathways. The first gene that was identified to play an essential role in determining the sex of the gonad was the Y-chromosome-encoded putative transcription factor Sry. In mice the Sry gene product has been shown to be both necessary and sufficient to impose a testis fate on the bipotential gonad (Lovell-Badge & Robertson 1990, Koopman et al. 1991). Although a recent report described a role for SRY in pre-mRNA splicing (Ohe et al. 2002), downstream targets of this master sex-determining gene remain elusive. Our understanding of sex determination and gonadal organo-genesis therefore have been dependent on targeted mutagenesis of factors that have been implicated to regulate Sry or to affect gonad formation in other model organisms or through spontaneous mutations leading to sex-reversal phenotypes in humans and mice (reviewed in Brennan & Capel 2004). Among these, the X-linked gene encoding the orphan nuclear receptor DAX1 (NR0B1, nuclear receptor subfamily 0, group B, member 1; Mouse Genome Informatics) was initially believed to cause XY sex reversal in mice when over-expressed in a Sry hypomorphic mutant background (Swain et al. 1998). Since over-expression of DAX1 led to the development of female reproductive structures from a bipotential gonad, albeit on a sensitized background, it was assumed that this gene was critical for ovarian development. Intriguingly, targeted deletion of Dax1, employing a conditional allele, does not lead to failure of ovarian development; on the contrary, postnatal spermatogenesis is impaired while null females are fully fertile (Yu et al. 1998). Furthermore, when the Dax1 ^+/– females were mated to males with the hypomorphic Sry allele, all XY offspring were sex-reversed and their external and internal genitalia were completely indistinguishable from their wild-type XX siblings (Meeks et al. 2003), suggesting that this gene interacts with Sry to determine gonadal sex in mice and is required not for ovarian development but for spermatogenesis, probably through its action in Sertoli cells.

It is critical to note, however, that the role of DAX1 is highly dependent on the genetic background of the mice under investigation. A recent study by Eicher and colleagues demonstrated that although DAX1 function is essential on the C57BL/6 (B6) background for testis development in XY mice, it is completely dispensable on the DBA/2J background (Bouma et al. 2005). In B6 Dax1 ^–/Y mice, there was a failure of up-regulation of Sox9, which is another important gene in the sex-determination pathway (see below), even though Sry expression initiated normally.

The homeobox gene Sox9 is the earliest marker of pre-Sertoli cells following Sry expression. Through the use of a conditional knockout model SOX9 has been shown to be essential for normal testis differentiation and sex-cord formation in the testis (Chaboissier et al. 2004). SOX9 regulates the differentiation of Sertoli cells and the production of male-specific markers Amh (anti-Müllerian hormone) and P450scc (cytochrome P450 side-chain cleavage), both of which are decreased in homozygous knockout mice. Indeed, this finding is a corroboration of earlier studies in which targeted disruption of the SOX9-binding site in the mouse Amh promoter caused a failure of transcription of Amh (see Table 2) that resulted in non-regression of the Müllerian duct structures in XY mice and the formation of pseudo-hermaphrodites, confirming that SOX9-mediated up-regulation of AMH is a critical and specific step that is essential in the male sex-determination pathway (Arango et al. 1999). Precisely how AMH mediates regression of Müllerian duct structures in vivo has been characterized by the generation of a conditional knockout for one of the type I BMP receptors, Bmpr1a (Jamin et al. 2002); male mice lacking these receptors in mesenchymal cells of the Müllerian duct fail to respond to AMH and have persistent uteri and oviducts.

Another gene involved in ovary development is Wnt4, which is required for Müllerian duct formation in both sexes prior to sex determination (Vainio et al. 1999); female mice lacking Wnt4 exhibit partial female–male sex reversal. Since Fst-null ovaries have a similar defect and since Fst expression is absent in Wnt4-null XX gonads, Fst, operating downstream of Wnt4, has been implicated in the suppression of BMP and/or activin signaling to suppress the male-specific coelomic vessel (Yao et al. 2004).

	Molecular control of pre-antral folliculogenesis

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Folliculogenesis in the mammalian ovary is one of the most intricately regulated developmental processes in biology. Not only does normal folliculogenesis entail the complex interplay of both long-range endocrine cues and short-range paracrine and autocrine factors, the development of viable offspring from fertilization depends, to a large extent, on the structural and functional integrity of the mature oocyte (Matzuk & Lamb 2002, Matzuk et al. 2002). Physiologically, it is useful to visualize ovarian follicles as divided into two broad groups, each being characterized by the presence or absence of a morphological antrum (cavity). Follicles prior to the formation of an antrum are independent of the pituitary gonadotropins for growth and survival; these pre-antral follicles undergo an extensive growth phase once they have been recruited into the active pool although they remain arrested in a prolonged diplotene stage of meiosis throughout this phase. In contrast, around the period of antrum formation, oocytes regain meiotic competence and oocytes in early antral follicles depend on pituitary follicle stimulating hormone (FSH) for survival and growth.

The formation of primordial follicles in the newborn mouse ovary is preceded by the development of germ-cell clusters in the embryonic ovary in which the dividing germ cells undergo incomplete cytokinesis to remain connected to each other through intercellular bridges and their subsequent breakdown perinatally (Pepling & Spradling 2001). The search for active players that regulate the formation of primordial follicles in mammals has resulted in the discovery of a bHLH (basic helix-loop-helix) transcription factor called FIG (factor in the germ line ; Figla) present specifically in the female germ cells. Mice carrying a targeted disruption of the Figla gene (Figla ^–/–) had comparable number of germ cells in their ovaries with respect to wild-type ovaries at the E 18–19 stage (Soyal et al. 2000). However, by postnatal day 1 (PD 1) the oocytes in the Figla ^–/– ovaries failed to grow in size and began to disappear; by PD 2 the null ovaries were completely devoid of primordial follicles. The Figla ^–/– oocytes do progress through meiosis I, but fewer cells survive beyond the diplotene stage, suggesting that FIG regulates the expression of survival factors/growth factors at this stage.

While Figla represents an oocyte-derived factor critical for the formation of primordial follicles, data from another mutant provide evidence that primordial follicle formation is regulated by the somatic components as well. Mutants lacking the Wnt4 gene product die shortly after birth due to defects in kidney development; however, at day 1 the mutant ovaries have a dramatic reduction in the number of meiotic-stage oocytes and most oocytes degenerate shortly thereafter (Vainio et al. 1999). A similar finding has emerged from mice with follistatin-deficient ovaries (Yao et al. 2004), highlighting the functional interactions between these two proteins (see above).

Shortly after birth a subset of primordial follicles is selected from a resting pool of quiescent follicles to undergo the follicular activation and growth. Follicular activation is characterized by oocyte growth and the transition of squamous to cuboidal granulosa cells, which signals the formation of a primary follicle followed by granulosa cell proliferation. Work from our group and others has recently identified several proteins believed to be critical at this step. A homeobox transcription factor, NOBOX, which is specific to the oocyte in the newborn ovary (Rajkovic et al. 2004), is essential in this process. In Nobox ^–/– ovaries, primordial follicles do not develop into primary follicles and beyond; instead, by PD 14, the majority of the primordial follicles are lost leading to ovarian failure and infertility. Future studies will illustrate the target genes for this transcription factor and uncover other oocyte-specific genes that play critical roles in this process.

Primordial follicles are recruited in cohorts for growth, differentiation, and maturation through a series of highly regulated processes. However, primordial follicles in Foxo3a ^–/– ovaries undergo global activation so that by PD 14 the null ovaries do not have any quiescent primordial follicles but have an almost 2-fold increase in the number of primary and secondary follicles as compared with wild-type ovaries. Significantly, the null ovaries also have an almost 4-fold increase in atretic follicles and an approx. 14-fold increase in the amount of zona pellucida remnants, suggesting that the majority of the growing follicles undergo atresia following the formation of the zona pellucida by the oocyte. Not surprisingly, therefore, this sequence of global activation and atresia leads to a premature depletion of the pool of primordial follicles and secondary infertility by 15 weeks of age (Castrillon et al. 2003, Hosaka et al. 2004).

Mutations in the human forkhead transcription factor FOXL2 cause an autosomal dominant condition referred to as Blepharophimosis/ptosis/epicanthus inversus syndrome (BPES), associated with premature ovarian failure (Crisponi et al. 2001, De Baere et al. 2001). Mice with a targeted disruption in the murine ortholog (Foxl2 ^–/– mice) of the human gene are infertile and provide a mechanism to explain the etiology of BPES. In Foxl2 ^–/– mice granulosa cells in the mutant ovaries do not make the squamouscuboidal transition and consequently the mutants lack any secondary follicles and are infertile (Schmidt et al. 2004). Another granulosacell derived factor that regulates the rate of follicle recruitment, in a subtler manner, is the anti-Müllerian hormone (AMH), which, when secreted from the granulosa cells of secondary follicles, acts in a paracrine fashion within the follicle to inhibit the recruitment of primordial follicles. In Amh ^–/– females there is a derepression of the follicular activation process causing mutant ovaries to be depleted of primordial follicles earlier than in wild-type ovaries (Durlinger et al. 1999).

Growth differentiation factor 9 (Gdf9) is an oocyte-derived mitogenic paracrine factor that plays a critical role in folliculogenesis by regulating the somatic components of the follicle. Gdf9 encodes for a member of the TGFß superfamily of secreted proteins that play important roles in cell proliferation and differentiation in mammals (Chang et al. 2002). Members of this family, such as GDF9, BMP15, AMH, and activins, exquisitely regulate early stages of folliculogenesis. In mice, a null mutation in Gdf9 causes complete infertility in female mice (Dong et al. 1996). Histopathological examination of null ovaries demonstrate a total lack of normal appearing secondary follicles indicating that GDF9 secreted from oocytes acts upon the granulosa cell population to augment the proliferation/differentiation of these follicular supporting cells. Thus, GDF9 is essential for the primary-to-secondary follicle transition in mice. The oocyte that has already entered the growth phase following activation continues to grow in these follicles (Carabatsos et al. 1998); however, devoid of the close support from the granulosa cell mass there is atresia of both the oocyte and eventually the somatic cells.

Another member of the TGFß superfamily of proteins that appears to play an important role in pre-antral stages of folliculogenesis is BMP15. BMP15 is secreted from oocytes and regulates the proliferation and differentiation of the granulosa cell population. Analysis of a null mutation in Bmp15 generated in our laboratory revealed that Bmp15 ^–/– mice are subfertile and display defects in ovulation and fertilization (Yan et al. 2001). However, studies in sheep have shown that BMP15 may also play important roles at the primary to secondary follicle transition (Galloway et al. 2000) or, alternatively, the mutant BMP15 may antagonize the functions of GDF9 (Liao et al. 2004).

	Spermatogenesis

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
In adult mammals, spermatogenesis is one of the most highly productive biological processes; one study (Johnson et al. 1992) done on human testis specimens obtained at autopsy estimated that the daily sperm production ranges between 3.5 and 7.5 million sperm/g of testis parenchyma. This tremendous capacity to produce spermatozoa throughout the life of an individual rests solely on the self-renewing capacity of spermatogonial stem cells, which constitute a small group of pluripotent cells present in the mature testis. Concomitant with the identification of molecular markers that identify the spermatogonial stem-cell pool (Kubota et al. 2004) has come new knowledge on the regulation, survival and proliferation of these rare populations of adult stem cells in vivo through knockout studies. Here we briefly review some of these models that have illustrated the utility of the targeted knockout approach.

Among the first mutations that were identified to have a defect in spermatogonia were the W and Sl mutations (Russell et al. 1979) that affect the Kit receptor, which is present on differentiating spermatogonia and its ligand, the Stem cell factor, which is produced in the Sertoli cells. However, mutations in the Kit signaling system affect only the differentiated spermatogonia as the receptor Kit is not expressed on earlier spermatogonial populations (Donovan & de Miguel 2001). Recently, glial cell-derived neurotrophic factor (GDNF), another Sertoli cell-derived factor distantly related to the TGFß family of growth factors, was shown, in a dose-dependent manner, to control the cell-fate decision of undifferentiated spermatogonia; reduced levels of GDNF caused spermatogonial differentiation and depletion of the stem-cell pool while high levels in vivo caused a differentiation block and increase in the numbers of spermatogonia (Meng et al. 2000). Germ-cell factors that play equally important roles in the proliferation of undifferentiated spermatogonia have also been identified through mouse knockouts; prominent has been the identification of a zinc-finger-containing transcriptional repressor, Plzf, that is essential for stem-cell self-renewal in the adult testis (Buaas et al. 2004, Costoya et al. 2004). Although the precise genes that are regulated by this transcription factor are not yet known, future studies utilizing transcriptome analysis of spermatogonia similar to the approach used by Costoya and colleagues (2004) should lead to further insights into the molecular mechanisms of germ-line stem-cell biology.

Knockouts have been created that identify essential genes at almost all steps of spermatogenesis (reviewed in Matzuk & Lamb 2002, de Rooij & de Boer 2003); literally hundreds of mutant mice have been characterized with arrest at various stages such as a spermatogonial differentiation block, pachytene stage meiotic arrest, and defects in spermiogenesis and sperm function.

Several genes that are expressed in post-meiotic cells have been found to be essential for spermiogenesis in a dose-sensitive manner. Notable among these are mutations in the genes encoding protamines 1 and 2 (Prm1 and Prm2), Kelch-like 10 (Klhl10) and sperm antigen 16 (Spag16), haploinsufficiency of which are incompatible with male fertility (Cho et al. 2001, Yan et al. 2004, Zhang et al. 2004). Male mice that are high-percentage chimeras of these genes are unable to transmit the mutant alleles through the germ line due to either severe morphological defects (as in the Prm1 or Prm2 mice) or severe reduction of sperm count and defects in sperm morphology and function (Klhl10 or Spag16). During spermiogenesis, chromatin packaging in the sperm head requires the replacement of histones by the transition proteins and protamines. Similar to protamines, deficiency of both transition proteins 1 and 2 also leads to severe defects in sperm morphology and infertility (Zhao et al. 2004), thereby highlighting the importance of proper chromatin remodeling during spermatogenesis.

Defects in sperm motility also lead to infertility; analysis of knockouts of CatSper1 (Ren et al. 2001) and CatSper2 (Quill et al. 2003), which are cationic channels present on the sperm membrane, a sperm-specific isoform of a glycolytic enzyme, Gapdhs (Miki et al. 2004), and of structural components of the axoneme (Spag6, Tekt2) have revealed interesting models of asthenozoospermia (Sapiro et al. 2002, Tanaka et al. 2004) that lead to infertility in male mice. Finally, knockouts have uncovered genes that play essential functions in the final step of sperm function, sperm–egg binding (reviewed in Evans 2001, Primakoff & Myles 2002). The plethora of single-knockout models that impair spermatogenesis at so many discrete steps is in addition to numerous models that cause a hitherto undefined or non-specific block in spermatogenesis and as a whole serve to illustrate both the complexity of this process and the huge number of potential targets for development of contraceptives in humans.

	Endocrine, paracrine and autocrine regulation of gametogenesis

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Germ cells are dependent on the somatic cells of the gonad (Sertoli and Leydig cells in the testis; granulosa and theca cells in the ovary) for development, which are in turn regulated by gonadotropins secreted by the pituitary, namely FSH and luteinizing hormone (LH). Deficiency of the pituitary gonadotropins leads to hypogonadism in both sexes, as in the hpg (hypogonadal) mouse, which carries a naturally occurring deletion in the gonadotropin-releasing hormone (GnRH) gene and consequently are infertile with reduction in the size of the gonads and accessory sex organs as well as a block in gametogenesis (Mason et al. 1986). Since hpg mice lack both LH and FSH, the relative contributions of each of these gonadotropin hormones to the phenotypic spectrum was not clear until our group and others created knockouts of four different genes in mice: Fshb (FSHß subunit), Fshr (FSH receptor), Lhb (LHß subunit) and Lhr (LH receptor). Analyzing the effects of these single knockouts revealed that FSH and LH had non-redundant roles in both sexes. In males, loss of FSH signaling resulted in smaller testis size, reduced spermatogenesis, and impaired sperm motility although the males were fertile (Kumar et al. 1997, Dierich et al. 1998). In contrast, females deficient for a functional FSH signaling pathway are sterile with thin uteri and hypoplastic ovaries due to a block in folliculogenesis prior to antrum formation (Kumar et al. 1997). An intact LH signaling pathway, however, is essential in both sexes as both males and females deficient for either LHß subunit or its cognate receptor develop hypogonadism and infertility (Lei et al. 2001, Zhang et al. 2001, Ma et al. 2004). Furthermore, folliculogenesis is blocked in LH-deficient ovaries at an early antral stage, suggesting that the progression from secondary to early antral follicle stage is dependent on FSH while LH is required for further development. Similarly in males, spermatogenesis is blocked at the round spermatid stage, revealing an absolute requirement for only LH in further maturation of the haploid gametic cells.

Sex hormones (estrogen and progesterone in females, testosterone in males) have been long believed to play essential roles in the development, differentiation and proper function of female and male reproductive systems. Targeted knockouts of their respective receptors have revealed precisely how these steroid hormones function and have uncovered hitherto unsuspected roles for estrogen signaling in males and testosterone in females. Estrogen acts through two nuclear hormone receptors, ER and ERß ; whereas deletion of ERß leads to subfertility (Krege et al. 1998) due to defects in granulosa cell development (Dupont et al. 2000), disruption of ER results in female infertility due to a complete lack of ovulation with the null ovaries displaying hemorrhagic cysts (Lubahn et al. 1993, Dupont et al. 2000). Interestingly, ER-knockout males are infertile as well due to a defect in fluid reabsorption in the epididymis that leads to degeneration of seminiferous tubules and disruption of spermatogenesis (Eddy et al. 1996, Hess et al. 1997), thereby identifying a male-specific role for estrogen. Progesterone, which also acts through two receptor isoforms (PR-A and PR-B), was believed to be important for the establishment and maintenance of the pregnant state; however, deletion of the progesterone receptor in females led to infertility due to defects in ovulation with associated uterine inflammation and disruption of mammary gland morphogenesis (Lydon et al. 1995). Identifying which of these effects were mediated by the two receptor isoforms required some subtle manipulations because both isoforms were generated from the same gene from alternative promoters and translation initiation sites. In elegant experiments, Mulac-Jericevic and colleagues (Mulac-Jericevic et al. 2000, 2003) mutated the two AUG initiation codons in embryonic stem cells and generated PR-A and PR-B knockouts (Table 2). Analysis of these mutants revealed essential roles for PR-A in ovulation and uterine physiology while PR-B was important for mammary gland development.

Studies from the spontaneous tfm (testicular feminized) mutant and from androgen receptor-knockout mice (Yeh et al. 2002) have shown androgen signaling to be essential for the descent of testes, development of male accessory sex organs, and spermatogenesis beyond the pachytene spermatocyte stage. Recently, generation of Sertoli cell-specific knockout of the androgen receptor have clarified the roles of testosterone signaling by demarcating androgen signaling in these cells as essential for spermatogenesis but not for non-gonadal functions of testosterone (Chang et al. 2004, De Gendt et al. 2004). Interestingly, signaling through the androgen receptor has also been found to have important roles in female fertility as null females show longer estrous cycles and reduced fertility (Hu et al. 2004).

Studies from our group and several others have also discovered the essential roles played by multiple TGFß family members in reproduction (reviewed by Pangas & Matzuk 2004, Jorgez et al. 2005; see Fig. 1). Among these, the Inha mouse model revealed unexpectedly important roles for the protein in gonadal cell-cycle regulation as null animals developed gonadal tumors of granulosa cell/Sertoli cell origin (Matzuk et al. 1992). As identified by our group and others, also important in reproduction are the activin signaling components; activin receptor type II mutants develop testicular hypoplasia and activin ß B subunit mutants have defective parturition (Vassalli et al. 1994, Matzuk et al. 1995). In order to analyze the functional interaction between activin ß A and activin ß B, our group also created a knockin allele of ß B into the ß A locus (Table 2). The knockin rescued the neonatal lethality of the ß A-knockout mice, thus allowing dissection of the reproductive phenotype (Brown et al. 2000). While male knockin mice had delayed fertility, female mice had severely reduced fertility, thus highlighting the important roles played by activin ß A in mammalian reproduction. Also significant is the follistatin conditional knockout model in which the activin/BMP/GDF antagonist follistatin is deleted only in the granulosa cells of the ovary. Female mice with a granulosa cell-specific deletion of follistatin have severely reduced fertility due to reduced numbers of ovarian follicles and defects in ovulation and fertilization (Jorgez et al. 2004). These and other models (reviewed by Chang et al. 2002) have identified an enormous number of proteins belonging to the TGFß superfamily with essential roles in reproduction.

View larger version (53K): [in this window] [in a new window]   Figure 1 TGFß superfamily signaling in germ-cell formation, sex determination, and ovarian folliculogenesis. Null mutations in several members of this family result in an arrest in the development of the reproductive system. Whereas Smad5 –/– embryos lack PGCs at E 8.0 (Chang & Matzuk 2001), Fst –/– mice have a loss of germ cells at E 18.5 ((Yao et al. 2004). Oocytes from Bmp15 –/–/Gdf9 +/– double mutants display defective cumulus-cell adherence to the oocytes (Yan et al. 2001). WT, wild type; ME, medulla; CO, cortex; ACVR2, activin receptor type 2.

	Early embryo: the maternal effect

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

	Conditional mutagenesis

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Although a knockout mouse is a valuable resource to analyze the effects of a null mutation in a tissue of interest, in a large number of cases the study of the phenotype is precluded by a more widespread effect caused by the mutation (for example, embryonic or perinatal lethality of homozygotes). Also, distinguishing between the cell-autonomous and cell-non-autonomous nature of action of a gene of interest is not possible in a ubiquitous knockout. Researchers have long recognized this drawback of the traditional knockout approach and have sought to develop technologies that would allow ablation of a gene function in a specific tissue and at a specific time: this is the rationale behind the conditional mutagenesis approach. Over the years refinement and innovation have led to a whole set of new tools with which conditional knockouts are achieved in vivo at present (see Lewandoski 2001 for a detailed review); here we discuss the advantages and pitfalls of the use of the binary Cre-loxP technology in the generation of conditional mutants within the context of reproduction. For a list of available Cre lines, go to www.mshri.on.ca/nagy/Cre-pub.html and see Table 1; for a list of available floxed alleles, go to www.mshri.on.ca/nagy/floxed.html.

The binary Cre-loxP method is so designed that the bacteriophage recombinase Cre is driven by a tissue- or stage-specific promoter in a restricted spatio-temporal manner; the deletion of gene function in that tissue is then dependent on the Cre recombinase to bind to its cognate recognition sequences (loxP) sites and effect a DNA recombination event between any two of such loxP sites, which can lead to deletion, inversion and other genomic rearrangements based on the position and orientation of such sites. The gene (or more commonly, the gene region) to be deleted is flanked by loxP (‘floxed’) sites in a targeting event similar to a conventional knockout; floxed mice are then bred with mice expressing Cre under the regulation of a promoter of choice to obtain tissue-specific knockouts in the offspring. Using this method several conditional knockouts have been generated that display reproductive defects (see Table 1); indeed, some of them were created after conventional knockouts resulted in embryonic or post-natal lethality (Oct4 and Fst).

One apparent deficit of this technology in reproduction is the paucity of available Cre- driver lines that are specifically expressed in reproductive tissues of interest. Another problem relates to the fact that our restricted knowledge of gene-regulatory elements results in use of promoter regions that prove to be ‘leaky’, meaning that they drive Cre expression and consequently gene deletion outside of their specific spatio-temporal zone (for example the Sf1-Cre line), thus defeating the very purpose of creating a conditional knockout. The last problem, and not a trivial one, is the issue of the efficiency of Cre-mediated recombination. Cre lines often express the recombinase at levels that are not sufficient to achieve complete recombination between loxP sites, thus resulting in the generation of a hypomorph. Attention has been focused on producing more efficient deleter Cre lines (Shimshek et al. 2002) but we are still some distance from achieving mastery over this elegant but at times difficult technology. Nevertheless, with the intensive efforts of various groups and the advent of newer technologies, for example inducible Cre lines, the Flp-Frt system, it is conceivable that in the future, instead of one cure for all ills, the field will witness the use of a multitude of technologies each best tailored in a particular situation to deliver the proper result: that of generation of a null mutation in the tissue of interest.

	In the future

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
Worldwide, 15% of couples are unable to conceive within 1 year of starting unprotected intercourse (Matzuk & Lamb 2002). Although fertility defects may be the secondary manifestation of systemic illnesses (endocrinopathies, neurological disorders, anatomic malformations), which may or may not be treatable, often the defect is primary to the reproductive system and requires specialized evaluation of urologists and/or gynecologists and fertility specialists for possible treatment. However, even with advanced clinical tools for diagnosis of reproductive defects, about a quarter of clinical infertility cases are diagnosed as idiopathic (Matzuk & Lamb 2002). Since the identification of the genetic bases of a complex disorder such as infertility cannot be done through traditional genetic approaches of pedigree analysis and linkage studies, our understanding of the genetics of reproduction in humans has remained minimal and is limited to a few documented examples such as the discovery of Y-chromosome microdeletions in azoospermic men or the findings of cystic fibrosis transmembrane conductance regulator (CFTR) mutations in patients with congenital bilateral absence of the vas deferens. This raises the possibility that some of these ‘idiopathic’ cases are likely due to as-yet unidentified genetic anomalies.

In this review we have illustrated some examples of how knockouts in mice have greatly enhanced our understanding of this complex biological process. This is not meant to be a comprehensive survey of all knockouts with reproductive defects but rather an illustrative essay of the utility of the knockout approach towards understanding a process. Challenges remain plenty in both our understanding of the process and in the use of this technology itself, but the emergence of exciting technologies and projects such as RNA interference in mammals (reviewed in Mittal 2004), the MICER project for genome-wide large-scale insertional mutagenesis from indexed targeting vectors (Adams et al. 2004), and the recent conceptualization of a whole-genome knockout mouse project, or KOMP, by an international collaboration (Austin et al. 2004) will broaden the scope and increase the enthusiasm of researchers in the power of knockouts and knockdowns in evaluating the roles of additional proteins in the reproductive axis in vivo.

	Acknowledgements

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 
The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.

	Footnotes

 
Received 12 July 2005
Accepted 12 September 2005

	References

Top Abstract Introduction Germ-line specification: what... PGC proliferation, migration and... Sex determination Molecular control of pre-antral... Spermatogenesis Endocrine, paracrine and... Early embryo: the maternal... Conditional mutagenesis In the future Acknowledgements References

 

Adams DJ, Biggs PJ, Cox T, Davies R, van der Weyden L, Jonkers J et al. 2004 Mutagenic insertion and chromosome engineering resource (MICER). Nature Genetics 36 867–871.[CrossRef][ISI][Medline]

Alfieri JA, Martin AD, Takeda J, Kondoh G, Myles DG & Primakoff P 2003 Infertility in female mice with an oocyte-specific knockout of GPI-anchored proteins. Journal of Cell Science 116 2149–2155.[Abstract/Free Full Text]

Ara T, Nakamura Y, Egawa T, Sugiyama T, Abe K, Kishimoto T et al. 2003 Impaired colonization of the gonads by primordial germ cells in mice lacking a chemokine, stromal cell-derived factor-1 (SDF-1). PNAS 100 5319–5323.[Abstract/Free Full Text]

Arango NA, Lovell-Badge R & Behringer RR 1999 Targeted mutagenesis of the endogenous mouse Mis gene promoter: in vivo definition of genetic pathways of vertebrate sexual development. Cell 99 409–419.[CrossRef][ISI][Medline]

Atchison FW, Capel B & Means AR 2003 Pin1 regulates the timing of mammalian primordial germ cell proliferation. Development 130 3579–3586.[Abstract/Free Full Text]

Austin CP, Battey JF, Bradley A, Bucan M, Capecchi M, Collins FS et al. 2004 The knockout mouse project. Nature Genetics 36 921–924.[CrossRef][ISI][Medline]

Bendel-Stenzel M, Anderson R, Heasman J & Wylie C 1998 The origin and migration of primordial germ cells in the mouse. Seminars in Cell Development and Biology 9 393–400.[CrossRef]

Bortvin A, Goodheart M, Liao M & Page DC 2004 Dppa3 /Pgc7 /stella is a maternal factor and is not required for germ cell specification in mice. BMC Developments in Biology 4 2.[CrossRef]

Bouma GJ, Albrecht KH, Washburn LL, Recknagel AK, Churchill GA & Eicher EM 2005 Gonadal sex reversal in mutant Dax1 XY mice: a failure to upregulate Sox9 in pre-Sertoli cells. Development 132 3045–3054.[Abstract/Free Full Text]

Bourc’his D, Xu GL, Lin CS, Bollman B & Bestor TH 2001 Dnmt3L and the establishment of maternal genomic imprints. Science 294 2536–2539.[Abstract/Free Full Text]

Bradley A, Evans M, Kaufman MH & Robertson E 1984 Formation of germ-line chimaeras from embryo-derived teratocarcinoma cell lines. Nature 309 255–256.[CrossRef][Medline]

Brennan J & Capel B 2004 One tissue, two fates: molecular genetic events that underlie testis versus ovary development. Nature Reviews in Genetics 5 509–521.[ISI][Medline]

Brown CW, Houston-Hawkins DE, Woodruff TK & Matzuk MM 2000 Insertion of Inhbb into the Inhba locus rescues the Inhba-null phenotype and reveals new activin functions. Nature Genetics 25 453–457.[CrossRef][ISI][Medline]

Buaas FW, Kirsh AL, Sharma M, McLean DJ, Morris JL, Griswold MD et al. 2004 Plzf is required in adult male germ cells for stem cell self-renewal. Nature Genetics 36 647–652.[CrossRef][ISI][Medline]

Burns KH, Viveiros MM, Ren Y, Wang P, DeMayo FJ, Frail DE et al. 2003 Roles of NPM2 in chromatin and nucleolar organization in oocytes and embryos. Science 300 633–636.[Abstract/Free Full Text]

Carabatsos MJ, Elvin J, Matzuk MM & Albertini DF 1998 Characterization of oocyte and follicle development in growth differentiation factor-9-deficient mice. Developmental Biology 204 373–384.[CrossRef][ISI][Medline]

Castrillon DH, Miao L, Kollipara R, Horner JW & DePinho RA 2003 Suppression of ovarian follicle activation in mice by the transcription factor Foxo3a. Science 301 215–218.[Abstract/Free Full Text]

Chaboissier MC, Kobayashi A, Vidal VI, Lutzkendorf S, van de Kant HJ, Wegner M et al. 2004 Functional analysis of Sox8 and Sox9 during sex determination in the mouse. Development 131 1891–1901.[Abstract/Free Full Text]

Chang C, Chen YT, Yeh SD, Xu Q, Wang RS, Guillou F et al. 2004 Infertility with defective spermatogenesis and hypotestosteronemia in male mice lacking the androgen receptor in Sertoli cells. PNAS 101 6876–6881.[Abstract/Free Full Text]

Chang H & Matzuk MM 2001 Smad5 is required for mouse primordial germ cell development. Mechanisms in Development 104 61–67.[CrossRef]

Chang H, Brown CW & Matzuk MM 2002 Genetic analysis of the mammalian transforming growth factor-beta superfamily. Endocrine Reviews 23 787–823.[Abstract/Free Full Text]

Cho C, Willis WD, Goulding EH, Jung-Ha H, Choi YC, Hecht NB & Eddy EM 2001 Haploinsufficiency of protamine-1 or -2 causes infertility in mice. Nature Genetics 28 82–86.[CrossRef][ISI][Medline]

Costoya JA, Hobbs RM, Barna M, Cattoretti G, Manova K, Sukhwani M et al. 2004 Essential role of Plzf in maintenance of spermatogonial stem cells. Nature Genetics 36 653–659.[CrossRef][ISI][Medline]

Crisponi L, Deiana M, Loi A, Chiappe F, Uda M, Amati P et al. 2001 The putative forkhead transcription factor FOXL2 is mutated in blepharophimosis/ptosis/epicanthus inversus syndrome. Nature Genetics 27 159–166.[CrossRef][ISI][Medline]

De Baere E, Dixon MJ, Small KW, Jabs EW, Leroy BP, Devriendt K et al. 2001 Spectrum of FOXL2 gene mutations in blepharophimosis-ptosis-epicanthus inversus (BPES) families demonstrates a genotype-phenotype correlation. Human Molecular Genetics 10 1591–1600.[Abstract/Free Full Text]

De Gendt K, Swinnen JV, Saunders PT, Schoonjans L, Dewerchin M, Devos A et al. 2004 A Sertoli cell-selective knockout of the androgen receptor causes spermatogenic arrest in meiosis. PNAS 101 1327–1332.[Abstract/Free Full Text]

de Rooij DG & de Boer P 2003 Specific arrests of spermatogenesis in genetically modified and mutant mice. Cytogenetics & Genome Research 103 267–276.[CrossRef]

de Sousa Lopes SM, Roelen BA, Monteiro RM, Emmens R, Lin HY, Li E et al. 2004 BMP signaling medi