BOR Papers in Press. Published on November 20, 2013 as DOI:10.1095/biolreprod.112.105809 1
Regulated Wnt/Beta-Catenin Signaling Sustains Adult Spermatogenesis in Mice1 Running title: Wnt signaling in the postnatal mouse testis Summary sentence: Contrasting models of disrupted Wnt signaling demonstrate its crucial role in normal adult spermatogenesis and facilitate identification of cell type specific-mediators of Wnt activity. Key words: Germ cells, spermatids, spermatocytes, testis, APC, mouse models Genevieve E. Kerr,4 Julia C. Young,4,5 Katja Horvay,4 Helen E. Abud,3,4 and Kate L. Loveland2,3,4,5 4 Department of Anatomy and Developmental Biology, Monash University, Clayton, Victoria Australia 5 Department of Biochemistry, Monash University, Clayton, Victoria Australia 1
Supported by National Health and Medical Research Council of Australia (#545916 to K.L. and #3236507 to K.L. and H.A.). 2 Correspondence: Kate L. Loveland. Departments of Anatomy & Developmental Biology and Biochemistry & Molecular Biology, Building 76 level 2, Monash University, Wellington Road, Clayton, Victoria 3168 Australia. E-mail: [email protected] 3 These authors contributed equally to this work. ABSTRACT The importance of Wnt signaling for postnatal testis function has been previously studied in several mouse models, with chronic pathway disruption addressing its function in Sertoli cells and in post-meiotic germ cells. While chronic beta-catenin deletion in Sertoli cells does not profoundly affect testis development, new data indicate that Wnt signaling is required at multiple stages of spermatogenesis. We used two mouse models which allow acute disruption of Wnt signaling to explore the importance of regulated Wnt pathway activity for normal germ cell development in adult male mice. Short term induction of mutations in Adenomatous polyposis coli (Apc) and beta-catenin (Ctnnbl), that increase and decrease Wnt signaling levels respectively, were generated in AhCre Apcfl/fl and AhCre Ctnnb1fl/fl mice. Each exhibited a distinct phenotype of disrupted spermatogenesis that was evident within 24 h and persisted for up to 4 days. Outcomes included germ cell apoptosis and rapid loss and altered blood testis barrier protein distribution and morphology. The functional significance of nuclear localized β-catenin protein in spermatocytes and round spermatids, indicative of active Wnt signaling, was highlighted by the profound loss of post-mitotic germ cells in both models. Developmentallyregulated Wnt signaling mediators identified through transcriptional profiling of wild type and AhCre Ctnnb1fl/fl mouse testes, identified Wnt receptors (e.g. Fzd4) and ligands (e.g. Wnt3, Wnt3a, Wnt5b, Wnt7a and Wnt8b). This demonstration that Wnt signaling control is essential for adult spermatogenesis supports the growing understanding that its disruption may underpin certain cases of male infertility.
Copyright 2013 by The Society for the Study of Reproduction.
2
INTRODUCTION Wnt signaling is a highly conserved pathway that is essential to diverse processes during development and in adult tissue homeostasis [1]. Perturbations of Wnt signaling can result in birth defects, cancer and other diseases [2]. The intracellular localization of β-catenin, the central mediator of canonical Wnt signaling, is considered to be indicative of Wnt pathway activation status [3, 4]. Its nuclear localization is regarded as a hallmark of active Wnt signaling due to the capacity of β-catenin to regulate target gene transcription, but β-catenin also has a role in cell adhesion at the plasma membrane [5, 6]. In the absence of a Wnt ligand, cytoplasmic β-catenin is sequestered by a “destruction complex” that contains adenomatous polyposis coli (APC), Axin, the protein kinase casein kinase 1 (CK1) and glycogen synthase kinase 3 (GSK3); the latter phosphorylates and targets β-catenin for proteasomal degradation [4]. Rapid pathway activation occurs when a Wnt ligand binds to membrane-associated receptor complexes comprised of Frizzled (Fzd) family members and a low-density lipoprotein receptor (LDLR)related protein (LRP). This causes dissociation of the destruction complex to enable first cytoplasmic accumulation of β-catenin, then its translocation into the nucleus to complex with Tcell factor/lymphoid enhancer binding factor (TCF/LEF) transcription factors to activate target genes (reviewed in [6]). Other Wnt-activated pathways that are independent of β-catenin, such as the planar cell polarity pathway, are described as non-canonical; these rely on distinct receptors and activate different signaling mediators, including JNKs [7]. Their activation has been shown to provoke alternative, and even contrasting, cellular responses to canonical Wnt signaling, for example in blood cell development [8]. However, many key developmental outcomes are driven by canonical Wnt signaling, and in our initial observations of mice in which this pathway was acutely disrupted for studies on gut stem cell biology [9], we noticed that male fertility was reduced. This outcome led to the present study examining the hypothesis that canonical Wnt signaling is integral to normal testis function in adults. The seminiferous epithelium of the mammalian testis constitutes a highly organised unit within which spermatogenesis is tightly regulated by interacting signaling networks of hormones and growth factors that are produced and processed by the surrounding epithelial and interstitial compartments [10-13]. During spermatogenesis, generations of spermatogonial stem cells, present at the basement membrane of the seminiferous tubules, undergo a progression of developmental steps involving mitosis, meiosis and differentiation to produce spermatozoa that are released into the seminiferous tubule lumen [14]. These processes are supported by the somatic Sertoli cells which surround the germ cells and form the epithelium essential for their normal maturation [12]. Each generation of germ cells develops in parallel with subsequent generations in a highly ordered, cyclic manner. These distinct and invariant cohorts are divided into 12 successive Stages of germ cell associations in mice which collectively represent the cycle of the adult seminiferous epithelium [14]. Wnt signaling is required for events in fetal life that underpin adult fertility. Wnt 4 is essential for normal male fetal reproductive tract development, acting downstream of Sry but upstream of Sox9 and desert hedgehog [15]. Wnt 4 mutant testes show compromised Sertoli cell differentiation, due to the action of Wnt 4 within the early genital ridge [15]. Wnt7a is similarly required, as its deletion inhibits Müllerian duct regression and causes retention of female reproductive tract tissues in adult males that can impede sperm passage at ejaculation [16]. Constitutive β-catenin activation in the Müllerian duct mesenchyme also leads to Müllerian duct
3
retention [17], highlighting the importance of balanced pathway signaling in fetal life for normal fertility. Several studies have recently established the contribution of regulated Wnt signaling within Sertoli cells to male fertility using mouse models that activate or block Wnt signaling (Table S1). Expression of a constitutively active β-catenin isoform in Sertoli cells beginning during embryogenesis under the control of AMH-Cre results in germ cell depletion by postnatal day 0, reduced levels of Sertoli cell identity markers, Sox9 and AMH, from embryonic day 13.5, and formation of Sertoli cell tumours after 8 months of age [18, 19]. Similarly, postnatal expression of activated β-catenin in Sertoli cells under the control of AMHR2-Cre results in infertility, of seminiferous tubule degeneration, progressive germ cell loss due to apoptosis and increased AMH expression [20, 21]. The activation of Wnt signaling via mutation of Apc also leads to impaired germ cell development and loss of Sertoli cell apical extensions and blood testis barrier integrity [22]. In contrast, deletion of β-catenin in Sertoli cells causes no phenotypic defects at embryonic or postnatal stages [18, 22, 23]. These studies collectively indicate that the constitutive activation of Wnt signaling in Sertoli cells causes spermatogenic defects, while deletion of β-catenin does not affect the capacity for Sertoli cells to support germline development. A limited number of studies have explored the importance of β-catenin selectively within germ cells. Expression of the stable, active form of β-catenin in primordial germ cell promoter TNAP-Cre-expressing cells and cell lineages results in male and female germ cell deficiency [24], while haploid germ cell β-catenin deletion (Prm1-Cre) causes reduced sperm count, increased germ cell apoptosis and impaired fertility, attributed by the study authors to the role of β-catenin in cell adhesion [25]. The role of Wnt signaling per se, however, was not examined in the latter study. The non-canonical ligand Wnt5a has been demonstrated to promote spermatogonial stem cell maintenance in culture independent of β-catenin, however, in vivo evidence of active Wnt signaling in spermatogonia has yet to be provided [26]. A murine stem cell spermatogonia cell-like line, C18-4, exhibited features of canonical Wnt signaling, responding to Wnt3a with proliferation and TOPFlash activities [27]. In post-mitotic germ cells, an inactivating mutation in naked cuticle 1, a Wnt pathway inhibitor, was associated with increased nuclear β-catenin in elongating spermatids and defective adult spermatogenesis [28]. These findings present an emerging picture in which Wnt ligands act at multiple sites to influence testis development and contribute to the functional integrity of the adult seminiferous epithelium. However, there is minimal information available concerning the identity of Wnt ligands within the testis, their sites of impact, and their specific influence on spermatogenic progression in vivo. Thus, while existing evidence shows that control of Wnt signaling levels is essential for Sertoli cell support of spermatogenesis, the absence of canonical Wnt signaling in Sertoli cells appears to have minimal effects on testis development [18, 22, 23] and the specific roles of Wnt signaling in germ cells remains to be defined. In the present study we show β-catenin-dependent Wnt signaling is active in germ cells of the adult testis and use two models of acutely altered Wnt signaling in which a profound impact is observed on germline cells following disruption to the normal levels of Wnt signaling. These findings identify particular germ cell types that are affected when canonical Wnt signaling is perturbed. Considered in light of our additional new data which identify the Wnt pathway components present in the developing and adult testis, we can provide a more robust understanding of the importance and mechanisms of Wnt signaling in the adult mouse testis.
4
MATERIALS AND METHODS Animals and Tissue Collection Wild type Swiss mice were obtained from Monash University Central Animal Services. Investigations conformed to the National Health and Medical Research Council/Commonwealth Scientific and Industrial Research Organisation/Australian Agricultural Council Code of Practice for the Care and Use of Animals for Experimental Purposes and were approved by the Monash Animal Research Platform Committee on Ethics in Animal Experimentation. Conditional deletion of β-catenin and Apc in the testes was achieved by breeding Ctnnb1fl/fl mice [29] (obtained from J. Huelsken, ISREC, Switzerland) and Apcfl/fl mice with mice containing the AhCre transgene (obtained from A. Clarke, Cardiff, United Kingdom) as previously described for intestinal studies [9, 30]. In this system, Cre recombinase is under the control of the cytochrome P4501A1 (CYP1A1) promoter element, and mice containing this transgene are referred to as AhCre mice [30]. Expression of Cre recombinase in AhCre mice is inducible by administration of β-napthoflavone (BNF) which binds to cytoplasmic aryl hydrocarbon receptors that are then transported into the nucleus where they bind DNA response elements in the CYP1A1 promoter and activate transcription [30, 31]. Cre recombinase expression is induced in the intestine and several other tissues following administration of BNF (80 mg per kg) to AhCre mice (Supplemental Figure S1; all Supplemental data is available online at www.biolreprod.org) [30]. Aryl hydrocarbon receptors that can efficiently bind BNF are present in mice from C57 and CBA genetic backgrounds [32] in many tissues including the seminiferous tubules and interstitial cells of the testis [33]. Control and test mice at 7 weeks of age were given daily intraperitoneal injections of BNF, and animals were killed for collection of testes on the following day; the number of injections varied between one and four. Testes were fixed immediately after collection in Bouins fixative for 5 hours, dehydrated through a graded ethanol series (70% to100%), embedded in paraffin, and sectioned at 5 μm onto Superfrost Plus 2 slides (Lomb Scientific, Sydney, Australia). Control samples, bearing one floxed allele of the gene, were treated simultaneously with test samples throughout. Histological Analysis Loss of germ cells by 5 days after BNF injection was measured for mice of both mutants strains by analysis of H&E sections. Three 5 μm sections, separated from each other by at least 150 μm were chosen from one testis per animal. Scans of whole testis sections were captured using an Aperio Scanscope XT (Aperio Technologies, Vista, California, USA) and analysed using ImageScope software (Aperio Technologies, California, USA) by placing a grid over the section which created lines 500 μm apart. Tubules touching these lines were evaluated. The most mature germ cell type present in each tubule cross-section was recorded using standard histological criteria [34]. Ninety to 550 tubules were scored per animal, and the proportion of all tubules for each animal containing elongating spermatids, round spermatids, pachytene spermatocytes, leptotene/zygotene spermatocytes, spermatogonia or lacking germ cells (i.e. Sertoli cell only) was calculated. The incidence of cellular apoptosis within the seminiferous epithelium was quantified by determining the proportion of rounded tubule cross sections with any TUNEL positive cells. Between 50 and 250 tubules were counted per animal and 3-4 animals were analysed per genotype. To quantify differences in cell proliferation, anti-PCNAstained sections were examined. All tubule cross sections contained PCNA positive cells, but their distribution was altered by tissue shrinkage in samples exhibiting loss of more mature germ
5
cell types. Thus we recorded the proportion of each seminiferous tubule periphery that lacked PCNA positive cells in 20-50 rounded tubule cross sections per animal, from 4-5 animals per genotype [35]. The incidence of undifferentiated spermatogonia was quantified by counting the number of Plzf-positive cells per rounded tubule cross section. Ten-40 seminiferous tubules were counted per animal for 3-4 animals per genotype. A Student’s T test was performed to determine statistical significance between groups, with p
Regulated Wnt/Beta-Catenin Signaling Sustains Adult Spermatogenesis in Mice1 Running title: Wnt signaling in the postnatal mouse testis Summary sentence: Contrasting models of disrupted Wnt signaling demonstrate its crucial role in normal adult spermatogenesis and facilitate identification of cell type specific-mediators of Wnt activity. Key words: Germ cells, spermatids, spermatocytes, testis, APC, mouse models Genevieve E. Kerr,4 Julia C. Young,4,5 Katja Horvay,4 Helen E. Abud,3,4 and Kate L. Loveland2,3,4,5 4 Department of Anatomy and Developmental Biology, Monash University, Clayton, Victoria Australia 5 Department of Biochemistry, Monash University, Clayton, Victoria Australia 1
Supported by National Health and Medical Research Council of Australia (#545916 to K.L. and #3236507 to K.L. and H.A.). 2 Correspondence: Kate L. Loveland. Departments of Anatomy & Developmental Biology and Biochemistry & Molecular Biology, Building 76 level 2, Monash University, Wellington Road, Clayton, Victoria 3168 Australia. E-mail: [email protected] 3 These authors contributed equally to this work. ABSTRACT The importance of Wnt signaling for postnatal testis function has been previously studied in several mouse models, with chronic pathway disruption addressing its function in Sertoli cells and in post-meiotic germ cells. While chronic beta-catenin deletion in Sertoli cells does not profoundly affect testis development, new data indicate that Wnt signaling is required at multiple stages of spermatogenesis. We used two mouse models which allow acute disruption of Wnt signaling to explore the importance of regulated Wnt pathway activity for normal germ cell development in adult male mice. Short term induction of mutations in Adenomatous polyposis coli (Apc) and beta-catenin (Ctnnbl), that increase and decrease Wnt signaling levels respectively, were generated in AhCre Apcfl/fl and AhCre Ctnnb1fl/fl mice. Each exhibited a distinct phenotype of disrupted spermatogenesis that was evident within 24 h and persisted for up to 4 days. Outcomes included germ cell apoptosis and rapid loss and altered blood testis barrier protein distribution and morphology. The functional significance of nuclear localized β-catenin protein in spermatocytes and round spermatids, indicative of active Wnt signaling, was highlighted by the profound loss of post-mitotic germ cells in both models. Developmentallyregulated Wnt signaling mediators identified through transcriptional profiling of wild type and AhCre Ctnnb1fl/fl mouse testes, identified Wnt receptors (e.g. Fzd4) and ligands (e.g. Wnt3, Wnt3a, Wnt5b, Wnt7a and Wnt8b). This demonstration that Wnt signaling control is essential for adult spermatogenesis supports the growing understanding that its disruption may underpin certain cases of male infertility.
Copyright 2013 by The Society for the Study of Reproduction.
2
INTRODUCTION Wnt signaling is a highly conserved pathway that is essential to diverse processes during development and in adult tissue homeostasis [1]. Perturbations of Wnt signaling can result in birth defects, cancer and other diseases [2]. The intracellular localization of β-catenin, the central mediator of canonical Wnt signaling, is considered to be indicative of Wnt pathway activation status [3, 4]. Its nuclear localization is regarded as a hallmark of active Wnt signaling due to the capacity of β-catenin to regulate target gene transcription, but β-catenin also has a role in cell adhesion at the plasma membrane [5, 6]. In the absence of a Wnt ligand, cytoplasmic β-catenin is sequestered by a “destruction complex” that contains adenomatous polyposis coli (APC), Axin, the protein kinase casein kinase 1 (CK1) and glycogen synthase kinase 3 (GSK3); the latter phosphorylates and targets β-catenin for proteasomal degradation [4]. Rapid pathway activation occurs when a Wnt ligand binds to membrane-associated receptor complexes comprised of Frizzled (Fzd) family members and a low-density lipoprotein receptor (LDLR)related protein (LRP). This causes dissociation of the destruction complex to enable first cytoplasmic accumulation of β-catenin, then its translocation into the nucleus to complex with Tcell factor/lymphoid enhancer binding factor (TCF/LEF) transcription factors to activate target genes (reviewed in [6]). Other Wnt-activated pathways that are independent of β-catenin, such as the planar cell polarity pathway, are described as non-canonical; these rely on distinct receptors and activate different signaling mediators, including JNKs [7]. Their activation has been shown to provoke alternative, and even contrasting, cellular responses to canonical Wnt signaling, for example in blood cell development [8]. However, many key developmental outcomes are driven by canonical Wnt signaling, and in our initial observations of mice in which this pathway was acutely disrupted for studies on gut stem cell biology [9], we noticed that male fertility was reduced. This outcome led to the present study examining the hypothesis that canonical Wnt signaling is integral to normal testis function in adults. The seminiferous epithelium of the mammalian testis constitutes a highly organised unit within which spermatogenesis is tightly regulated by interacting signaling networks of hormones and growth factors that are produced and processed by the surrounding epithelial and interstitial compartments [10-13]. During spermatogenesis, generations of spermatogonial stem cells, present at the basement membrane of the seminiferous tubules, undergo a progression of developmental steps involving mitosis, meiosis and differentiation to produce spermatozoa that are released into the seminiferous tubule lumen [14]. These processes are supported by the somatic Sertoli cells which surround the germ cells and form the epithelium essential for their normal maturation [12]. Each generation of germ cells develops in parallel with subsequent generations in a highly ordered, cyclic manner. These distinct and invariant cohorts are divided into 12 successive Stages of germ cell associations in mice which collectively represent the cycle of the adult seminiferous epithelium [14]. Wnt signaling is required for events in fetal life that underpin adult fertility. Wnt 4 is essential for normal male fetal reproductive tract development, acting downstream of Sry but upstream of Sox9 and desert hedgehog [15]. Wnt 4 mutant testes show compromised Sertoli cell differentiation, due to the action of Wnt 4 within the early genital ridge [15]. Wnt7a is similarly required, as its deletion inhibits Müllerian duct regression and causes retention of female reproductive tract tissues in adult males that can impede sperm passage at ejaculation [16]. Constitutive β-catenin activation in the Müllerian duct mesenchyme also leads to Müllerian duct
3
retention [17], highlighting the importance of balanced pathway signaling in fetal life for normal fertility. Several studies have recently established the contribution of regulated Wnt signaling within Sertoli cells to male fertility using mouse models that activate or block Wnt signaling (Table S1). Expression of a constitutively active β-catenin isoform in Sertoli cells beginning during embryogenesis under the control of AMH-Cre results in germ cell depletion by postnatal day 0, reduced levels of Sertoli cell identity markers, Sox9 and AMH, from embryonic day 13.5, and formation of Sertoli cell tumours after 8 months of age [18, 19]. Similarly, postnatal expression of activated β-catenin in Sertoli cells under the control of AMHR2-Cre results in infertility, of seminiferous tubule degeneration, progressive germ cell loss due to apoptosis and increased AMH expression [20, 21]. The activation of Wnt signaling via mutation of Apc also leads to impaired germ cell development and loss of Sertoli cell apical extensions and blood testis barrier integrity [22]. In contrast, deletion of β-catenin in Sertoli cells causes no phenotypic defects at embryonic or postnatal stages [18, 22, 23]. These studies collectively indicate that the constitutive activation of Wnt signaling in Sertoli cells causes spermatogenic defects, while deletion of β-catenin does not affect the capacity for Sertoli cells to support germline development. A limited number of studies have explored the importance of β-catenin selectively within germ cells. Expression of the stable, active form of β-catenin in primordial germ cell promoter TNAP-Cre-expressing cells and cell lineages results in male and female germ cell deficiency [24], while haploid germ cell β-catenin deletion (Prm1-Cre) causes reduced sperm count, increased germ cell apoptosis and impaired fertility, attributed by the study authors to the role of β-catenin in cell adhesion [25]. The role of Wnt signaling per se, however, was not examined in the latter study. The non-canonical ligand Wnt5a has been demonstrated to promote spermatogonial stem cell maintenance in culture independent of β-catenin, however, in vivo evidence of active Wnt signaling in spermatogonia has yet to be provided [26]. A murine stem cell spermatogonia cell-like line, C18-4, exhibited features of canonical Wnt signaling, responding to Wnt3a with proliferation and TOPFlash activities [27]. In post-mitotic germ cells, an inactivating mutation in naked cuticle 1, a Wnt pathway inhibitor, was associated with increased nuclear β-catenin in elongating spermatids and defective adult spermatogenesis [28]. These findings present an emerging picture in which Wnt ligands act at multiple sites to influence testis development and contribute to the functional integrity of the adult seminiferous epithelium. However, there is minimal information available concerning the identity of Wnt ligands within the testis, their sites of impact, and their specific influence on spermatogenic progression in vivo. Thus, while existing evidence shows that control of Wnt signaling levels is essential for Sertoli cell support of spermatogenesis, the absence of canonical Wnt signaling in Sertoli cells appears to have minimal effects on testis development [18, 22, 23] and the specific roles of Wnt signaling in germ cells remains to be defined. In the present study we show β-catenin-dependent Wnt signaling is active in germ cells of the adult testis and use two models of acutely altered Wnt signaling in which a profound impact is observed on germline cells following disruption to the normal levels of Wnt signaling. These findings identify particular germ cell types that are affected when canonical Wnt signaling is perturbed. Considered in light of our additional new data which identify the Wnt pathway components present in the developing and adult testis, we can provide a more robust understanding of the importance and mechanisms of Wnt signaling in the adult mouse testis.
4
MATERIALS AND METHODS Animals and Tissue Collection Wild type Swiss mice were obtained from Monash University Central Animal Services. Investigations conformed to the National Health and Medical Research Council/Commonwealth Scientific and Industrial Research Organisation/Australian Agricultural Council Code of Practice for the Care and Use of Animals for Experimental Purposes and were approved by the Monash Animal Research Platform Committee on Ethics in Animal Experimentation. Conditional deletion of β-catenin and Apc in the testes was achieved by breeding Ctnnb1fl/fl mice [29] (obtained from J. Huelsken, ISREC, Switzerland) and Apcfl/fl mice with mice containing the AhCre transgene (obtained from A. Clarke, Cardiff, United Kingdom) as previously described for intestinal studies [9, 30]. In this system, Cre recombinase is under the control of the cytochrome P4501A1 (CYP1A1) promoter element, and mice containing this transgene are referred to as AhCre mice [30]. Expression of Cre recombinase in AhCre mice is inducible by administration of β-napthoflavone (BNF) which binds to cytoplasmic aryl hydrocarbon receptors that are then transported into the nucleus where they bind DNA response elements in the CYP1A1 promoter and activate transcription [30, 31]. Cre recombinase expression is induced in the intestine and several other tissues following administration of BNF (80 mg per kg) to AhCre mice (Supplemental Figure S1; all Supplemental data is available online at www.biolreprod.org) [30]. Aryl hydrocarbon receptors that can efficiently bind BNF are present in mice from C57 and CBA genetic backgrounds [32] in many tissues including the seminiferous tubules and interstitial cells of the testis [33]. Control and test mice at 7 weeks of age were given daily intraperitoneal injections of BNF, and animals were killed for collection of testes on the following day; the number of injections varied between one and four. Testes were fixed immediately after collection in Bouins fixative for 5 hours, dehydrated through a graded ethanol series (70% to100%), embedded in paraffin, and sectioned at 5 μm onto Superfrost Plus 2 slides (Lomb Scientific, Sydney, Australia). Control samples, bearing one floxed allele of the gene, were treated simultaneously with test samples throughout. Histological Analysis Loss of germ cells by 5 days after BNF injection was measured for mice of both mutants strains by analysis of H&E sections. Three 5 μm sections, separated from each other by at least 150 μm were chosen from one testis per animal. Scans of whole testis sections were captured using an Aperio Scanscope XT (Aperio Technologies, Vista, California, USA) and analysed using ImageScope software (Aperio Technologies, California, USA) by placing a grid over the section which created lines 500 μm apart. Tubules touching these lines were evaluated. The most mature germ cell type present in each tubule cross-section was recorded using standard histological criteria [34]. Ninety to 550 tubules were scored per animal, and the proportion of all tubules for each animal containing elongating spermatids, round spermatids, pachytene spermatocytes, leptotene/zygotene spermatocytes, spermatogonia or lacking germ cells (i.e. Sertoli cell only) was calculated. The incidence of cellular apoptosis within the seminiferous epithelium was quantified by determining the proportion of rounded tubule cross sections with any TUNEL positive cells. Between 50 and 250 tubules were counted per animal and 3-4 animals were analysed per genotype. To quantify differences in cell proliferation, anti-PCNAstained sections were examined. All tubule cross sections contained PCNA positive cells, but their distribution was altered by tissue shrinkage in samples exhibiting loss of more mature germ
5
cell types. Thus we recorded the proportion of each seminiferous tubule periphery that lacked PCNA positive cells in 20-50 rounded tubule cross sections per animal, from 4-5 animals per genotype [35]. The incidence of undifferentiated spermatogonia was quantified by counting the number of Plzf-positive cells per rounded tubule cross section. Ten-40 seminiferous tubules were counted per animal for 3-4 animals per genotype. A Student’s T test was performed to determine statistical significance between groups, with p
