Prepubertal testis development relies on retinoic acid but not rexinoid receptors in Sertoli cells
2006; Springer Nature; Volume: 25; Issue: 24 Linguagem: Inglês
10.1038/sj.emboj.7601447
ISSN1460-2075
AutoresNadège Vernet, Christine Dennefeld, Florian Guillou, Pierre Chambon, Norbert B. Ghyselinck, Manuel Mark,
Tópico(s)Retinoids in leukemia and cellular processes
ResumoArticle23 November 2006free access Prepubertal testis development relies on retinoic acid but not rexinoid receptors in Sertoli cells Nadège Vernet Nadège Vernet IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Christine Dennefeld Christine Dennefeld IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Florian Guillou Florian Guillou INRA, Université de Tours, Nouzilly, France Search for more papers by this author Pierre Chambon Pierre Chambon IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Norbert B Ghyselinck Corresponding Author Norbert B Ghyselinck IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Manuel Mark Corresponding Author Manuel Mark IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Nadège Vernet Nadège Vernet IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Christine Dennefeld Christine Dennefeld IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Florian Guillou Florian Guillou INRA, Université de Tours, Nouzilly, France Search for more papers by this author Pierre Chambon Pierre Chambon IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Norbert B Ghyselinck Corresponding Author Norbert B Ghyselinck IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Manuel Mark Corresponding Author Manuel Mark IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France Search for more papers by this author Author Information Nadège Vernet1, Christine Dennefeld1, Florian Guillou2, Pierre Chambon1, Norbert B Ghyselinck 1 and Manuel Mark 1 1IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS, Faculté de Médecine, INSERM, U596, Illkirch, France, UMR7104, Illkirch, France, Strasbourg, France 2INRA, Université de Tours, Nouzilly, France *Corresponding authors: IGBMC, 1 rue Laurent Fries, BP10142, Illkirch F-67404, France. Tel.: +33 388 655 674; Fax: +33 388 653 201; E-mail: [email protected] or Tel.: +33 388 655 636; Fax: +33 388 653 201; E-mail: [email protected] The EMBO Journal (2006)25:5816-5825https://doi.org/10.1038/sj.emboj.7601447 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Sertoli cells (SC) are instrumental to stem spermatogonia differentiation, a process that critically depends on retinoic acid (RA). We show here that selective ablation of RA receptor alpha (RARalpha) gene in mouse SC, singly (RaraSer−/− mutation) or in combination with RARbeta and RARgamma genes (Rara/b/gSer−/− mutation), abolishes cyclical gene expression in these cells. It additionally induces testis degeneration and delays spermatogonial expression of Stra8, two hallmarks of RA deficiency. As identical defects are generated upon inactivation of RARalpha in the whole organism, our data demonstrate that all the functions exerted by RARalpha in male reproduction are Sertoli cell-autonomous. They further indicate that RARalpha is a master regulator of the cyclical activity of SC and controls paracrine pathways required for spermatogonia differentiation and germ cell survival. Most importantly, we show that the ablation of all RXR (alpha, beta and gamma isotypes) in SC does not recapitulate the phenotype generated upon ablation of all three RARs, thereby providing the first evidence that RARs exert functions in vivo independently of RXRs. Introduction The mammalian seminiferous epithelium consists of Sertoli cells (SC) and germ cells. Its development, renewal and functioning, which underlie spermatogenesis, require a complex assortment of hormones and cytokines (Meng et al, 2000; Holdcraft and Braun, 2004). Among these signals, retinoic acid (RA), the active metabolite of vitamin A (retinol), regulates spermatogonia differentiation and spermatid adhesion properties (Ghyselinck et al, 2006; Vernet et al, 2006). RA acts through binding to nuclear retinoic acid receptors (RARα, β and γ isotypes), which are ligand-dependent transcriptional regulators transducing the RA signal in the form of heterodimers with the rexinoid receptors, RXRα, β and γ (Kastner et al, 1997; Chambon, 2005; Mark et al, 2006). During post-natal development and adulthood, each RAR is detected predominantly in a specific cell type of the seminiferous epithelium: RARα in SC, RARβ in round spermatids and RARγ in A spermatogonia (Vernet et al, 2006). Germline inactivation of Rara results in testis degeneration comprising features observed upon dietary vitamin A deficiency (VAD), whereas those of Rarb or Rarg do not cause primary testis defects (Lufkin et al, 1993; Vernet et al, 2006). Stem spermatogonia have remarkable ability to both self-renew and differentiate. The balance between these two processes is thought to depend on a proper environment, which is provided by the supporting SC (Payne and Braun, 2006; Ryu et al, 2006, and references therein). In addition, SC are essential to initiate spermatogenesis at puberty, and to maintain it after sexual maturity (Sharpe et al, 2003). SC display cyclical changes in morphology, gene expression and biochemical activity (Morales and Clermont, 1993; Parvinen, 1993), which are associated with stages of the seminiferous epithelium cycle, a series of constant germ cell associations reflecting coordination of meiosis and spermiogenesis (i.e., spermatid maturation; Russell et al, 1990). As the cyclical activity of SC is established before any signs of heterogeneity in germ cell populations, it may be involved in initiating spermatogenesis at puberty (Timmons et al, 2002). Given the central role of SC in spermatogonial stem cell self-renewal and spermatogenesis, we were interested in generating SC-specific Rar and Rxr knockouts. Using this genetic approach, we demonstrate that RARα is cell-autonomously instrumental to the cyclical activity of SC and to the structural integrity of the seminiferous epithelium, in contrast to RXRs, which are dispensable. Results SC-specific ablation of RARα in mice Mice carrying loxP-flanked alleles of Rara (Chapellier et al, 2002a) were crossed with mice bearing the Amh-Cre transgene (Lecureuil et al, 2002) to generate RaraSer−/− mutants, in which both alleles of Rara were excised in SC. These crosses also generated control males carrying two loxP-flanked alleles of Rara, which did not display histological defects and are hereafter referred to as wild-type (WT) mice. Importantly, no immunostaining for RARα was detected in the testes of RaraSer−/− adult mice (Supplementary Figure 1), indicating that RaraSer−/− mutants actually lack RARα in SC. Note that, as the Amh–Cre transgene is expressed from embryonic day 15.5 onwards (Lecureuil et al, 2002), excision of Rara occurs before the onset of spermatogenesis at postnatal day 5 (P5) (Bellve et al, 1977). Ablation of RARα in SC results in reduced spermatogenesis and age-dependent testis degeneration In young, 9-week-old RaraSer−/− mutants (n=6), spermatogenesis yielded mature spermatids in which nuclear elongation as well as acrosome and flagellum development was complete (St16; Figure 1F and Supplementary Figure 2). However, these mature spermatids failed to align at the luminal side of the seminiferous epithelium (St16; compare Figure 1E with F), and were often not released. They were instead retained within the epithelium (St16r; compare Figure 1G and H). In addition, they were scarce (St16; compare Figure 1E and F), and frequently exhibited ultrastructural abnormalities indicative of necrosis (Supplementary Figure 2B). The seminiferous epithelium of RaraSer−/− mutants also showed large vacuoles (VA; compare Figure 1A and B), and desquamation of round spermatids (black arrowhead in Figure 1B and K, compare with 1A and J). In keeping with these defects, the caudal epididymis contained low spermatozoa stores (compare SZ in Figure 1C and D), but numerous round spermatids (R; Figure 1D); all these cells were necrotic (Supplementary Figure 2J). Altogether, these data are indicative of complete but reduced spermatogenesis and testis degeneration (Holstein et al, 2003). A decreased production and a failure of detachment of mature spermatids contribute to the reduced spermatogenesis in RaraSer−/− mutants. To investigate whether cell death also contributed to this phenotype, TUNEL assays were performed in 9-week-old testes (Table I). Numerous TUNEL-positive spermatocytes and round spermatids were detected in the mutant testes (compare Figure 1K with J), and TUNEL-positive elongated spermatids (yellow arrowheads, Figure 1L–N) were present in 68±1% (mean±s.e.m.; n=3 mutants) of the tubule sections versus 5±1% in WT testes (n=3). Thus, lack of RARα markedly impairs the SC capacity to support survival of meiotic (spermatocytes) and post-meiotic germ cells (spermatids). Figure 1.Ablation of RARα in SC yields a progressive testis degeneration resulting from spermatid desquamation and germ cell apoptosis, but not from a disruption of the seminiferous epithelium cycle. (A–I) Histological sections stained with hematoxylin and eosin (A–D, I) or toluidine blue (E–H) through the testes (A, B, E–I) or epididymides (C, D) of 9-week-old (A–H) and 12-month-old (I) mice. (J–N) TUNEL assays: note that in panels J, K and N, the positive signal was converted to a red false color and superimposed with the DAPI nuclear stain (blue false color). (O–T) Identification of the descendants of preleptotene spermatocytes, 18 days following a single injection of BrdU. (P) and (S) are superimpositions of the BrdU-labeled step 8 and 9 spermatids (red false color) with a periodic acid Schiff (PAS) counterstain; (Q) and (T) are high-magnification views of the boxed areas. L, leptotene spermatocytes; P, pachytene spermatocytes; Pr, preleptotene spermatocytes; R, round germ cells; St7, St9, St16, step 7, 9 and 16 spermatids, respectively; St16r, retained step 16 spermatids; SZ, spermatozoa; VA, vacuoles. Asterisks and double asterisks indicate tubules containing SC only and full complement of germ cells, respectively. The black and yellow arrowheads point to round spermatids detaching from the seminiferous epithelium and to TUNEL-positive elongated spermatids, respectively. Roman numerals designate stages of the seminiferous epithelium cycle. Bar (in T): 80 μm (A, B, I–K, O, P, R, S), 30 μm (C, D, Q, T), 20 μm (E–H) and 15 μm (L–N). Download figure Download PowerPoint Table 1. Percentage of tubule cross-sections containing TUNEL-positive cells in 9-week-old WT and RaraSer−/− mice, and distribution of apoptotic cell types WT RaraSer−/− Percentage of tubule sections containing TUNEL-positive round germ cells 18±1 58±12 Number of TUNEL-positive round germ cells in a testis cross-section 46±6 225±45 Percentage of spermatocytes 89±3 80±16 Percentage of spermatids 1±1 17±1 Percentage of unidentified 10±1 3±1 Mean±s.e.m.; n=3 in each group of age. The number of tubule sections analyzed in each testis was between 120 and 140. In 12-month-old RaraSer−/− mutants (n=3), up to 86% of tubule sections contained only SC (asterisks; Figure 1I and Supplementary Table I), indicating that (i) RARα in SC is also necessary for the survival of spermatogonia, and (ii) testis degeneration (i.e., vacuolation of the seminiferous epithelium, desquamation of immature spermatids and germ cell death) becomes more severe upon aging. Ablation of RARα in SC does not affect their polarity and their density At the ultrastructural level, SC of RaraSer−/− mutants were essentially normal and the blood testis barrier appeared unaffected (Supplementary Figure 2). As a given SC can support survival and differentiation of only a limited number of germ cells (Sharpe et al, 2003), we assumed that germ cell desquamation and apoptosis observed in RaraSer−/− mutants could be accounted for by a reduced SC density. However, no significant difference in SC density was noted between WT (26.7±0.8 cells/mm; mean±s.e.m.; n=40 tubule sections) and RaraSer−/− testes (28.7±1.1 cells/mm; n=44 tubule sections) at 9 weeks of age (Supplementary Figure 4), indicating that the ratio of SC to germ cells is normal in RaraSer−/− mutants. Ablation of RARα in SC does not affect the cycle of the seminiferous epithelium or the duration of spermatogenesis The different generations of germ cells, while synchronously progressing through spermatogenesis, form cellular associations of fixed composition (called epithelial stages) that follow each other according to a stereotyped sequence known as the seminiferous epithelium cycle. Twelve epithelial stages (I–XII) can be identified in the mouse (Russell et al, 1990). In 9-week-old RaraSer−/− testes, the 12 epithelial stages were readily identifiable and each occupied the circumference of a seminiferous tubule cross-section (Figure 1E–H). The duration of meiotic and post-meiotic phases of spermatogenesis was evaluated by identifying, 18 days after a single injection of BrdU, the labeled descendants of preleptotene spermatocytes. In both WT and RaraSer−/− mice, the most advanced BrdU-labeled cells were step 9 spermatids, and there was no labeling in germ cells younger than step 8 spermatids (Figure 1O–T). These data indicate that the absence of RARα in SC does not alter the seminiferous epithelium cycle or the duration of meiosis and spermiogenesis. Absence of RARα or of RA in SC delays the progression of the prepubertal wave of spermatogenesis To analyze the impact of the RaraSer−/− mutation on the prepubertal wave of spermatogenesis, which is normally completed by postnatal day 35 (P35), we compared development of the seminiferous epithelium in RaraSer−/− and WT males (n=3 for each genotype and age group) at P5 (i.e., when gonocytes differentiate into primitive A spermatogonia), P10 (i.e., at the onset of meiosis), P20 (i.e., when post-meiotic cells first appear), P25 and P30 (Bellve et al, 1977). At P5, RaraSer−/− and WT seminiferous cords were morphologically indistinguishable (not shown), although the cords expressing the spermatogonia differentiation marker Stra8 were significantly fewer in RaraSer−/− testes (Figure 2A and Supplementary Figure 5). At P10, leptotene spermatocytes, which represented the most advanced germ cell type, were also fewer in RaraSer−/− testes (Figure 2B). At P20, a majority of RaraSer−/− tubule sections did not contain spermatocytes beyond the zygotene stage, whereas the vast majority of their WT counterparts displayed more advanced pachytene and diplotene spermatocytes (Figure 2C). Along the same lines, post-meiotic round spermatids were absent in RaraSer−/− testes at P20, but were always present in age-matched WT testes (Figure 2C). These data are indicative of a delay in spermatogenesis, which interestingly is not related to testis degeneration as increase in germ cell apoptosis (Figure 2D) and vacuolation of the seminiferous epithelium (not shown) were not observed in RaraSer−/− testes before P20 and P25, respectively. Figure 2.Ablation of RA signaling in SC delays the first spermatogenic cycle without altering the normal timing of androgen receptor expression. (A) Percentage of seminiferous cord cross-sections containing Stra8-positive spermatogonia in WT (blue bars) and in RaraSer−/− (purple bars) testes at P5. (B, C) Percentages of seminiferous cord or tubule cross-sections in which leptotene spermatocytes (L), preleptotene/leptotene (PR+L), zygotene (Z), pachytene (P) and diplotene (D) spermatocytes and round spermatids (R) represent the most advanced germ cell types in WT (blue bars) and RaraSer−/− (purple bars) testes at P10 (B) and P20 (C). (D) Percentages of seminiferous cord or tubule cross-sections containing at least one (first set of bars) or at least three (second set of bars) apoptotic germ cell in WT (blue bars) and in RaraSer−/− (purple bars) testes at P10, P20 and P25. Note that in panels A–D, the bars represent mean±s.e.m. (n=3–5); the asterisks indicate a significant difference (*P<0.05; ***P<0.001). (E, F) Detection of Stra8 transcripts at P5 in WT and Aldh1a1-null testes: although histologically indistinguishable at this developmental stage, the seminiferous cord sections containing Stra8-positive spermatogonia are much less abundant in the Aldh1a1-null than in the WT testis. (G–J) Immunodetection of androgen receptor (red signal) at the onset of spermatogenesis (i.e., P5) and at the beginning of meiosis (i.e., P10). At P5, the androgen receptor is detected in all peritubular myoid cell precursors, as well as occasionally and weakly in immature SC. At P10, the androgen receptor is expressed in peritubular myoid cells and in all immature SC. G, spermatogonia; PR, L, Z, P, D, preleptotene, leptotene, zygotene, pachytene and diplotene spermatocytes, respectively; M, peritubular myoid cells. S, immature Sertoli cells; R, spermatids. Bar (in J): 200 μm (E, F) and 50 μm (G–J). Download figure Download PowerPoint Androgen and FSH signaling pathways in SC play essential functions in testis development, as inactivation of androgen and FSH receptors (Ar and Fshr, respectively) delay the prepubertal wave of spermatogenesis (Chang et al, 2004; De Gendt et al, 2004; Johnston et al, 2004). However, these two pathways are not involved in the delay of prepubertal spermatogenesis in RaraSer−/− mutants as the expression pattern of Ar (Figure 2G–J) and Fshr (Supplementary Figure 5) were normal in immature SC of RaraSer−/− testes. To investigate whether the delay in spermatogonia differentiation observed in testes lacking RARα could be mimicked upon decreasing RA availability, we analyzed expression of Stra8 in the testes of Aldh1a1-null mice lacking retinaldehyde dehydrogenase 1 (Matt et al, 2005), which is the main RA-synthesizing enzyme in SC at P5 (Vernet et al, 2006). In Aldh1a1-null testes at P5, only 4±3% (mean±s.e.m.; n=3) of seminiferous cords contained spermatogonia expressing Stra8 versus 28±3% (mean±s.e.m.; n=3) in WT littermates (compare Figure 2E with F). Therefore, a RA-liganded RARα in SC is required for proper spermatogonia differentiation during the prepubertal wave of spermatogenesis. Cyclical expression of numerous genes is lost in SC lacking RARα The cyclical activity of SC was investigated through analysis of genes known to display SC-restricted cyclical expression, such as the androgen receptor (AR) whose expression peaks at stage VI–VIII (Figure 3A–D; Zhou et al, 2002), and the membrane protein STRA6 expressed at stages VII–VIII and IX (Figure 4A and B; Bouillet et al, 1997). In RaraSer−/− adults, all SC displayed similar levels of AR (Figure 3E–H), and low, uniform, levels of STRA6 from stage I to stage XII (Figure 4C and D). Along these lines, the stage-dependent variations of expression of GATA1 (Yomogida et al, 1994), galectin-1 (Timmons et al, 2002), clusterin (Morales et al, 1987) and procathepsin L (Wright et al, 2003) were all lost in the RaraSer−/− seminiferous epithelium (Supplementary Figures 6 and 7). Altogether, these data indicate that the cyclical activity of SC is abolished in RaraSer−/− mutants. Importantly, the cyclical expression of Stra8 in spermatocytes and A spermatogonia (Figure 4E–H; Supplementary Figure 7) and Rxra in round spermatids (not shown) were not modified in RaraSer−/− testes. These observations are in keeping with our histological findings that ablation of RARα in SC does not alter the seminiferous epithelium cycle (Figure 1E–H). Figure 3.Ablation of RARα in SC abrogates the epithelial stage-dependent variations of AR expression. Immunohistochemical detection of AR in the seminiferous epithelium at 9 weeks of age. (A–D) In WT testis, immunolabeling for AR is strong in SC nuclei at stages VII and VIII and weak at other epithelial stages. (E–H) In RaraSer−/− mutant testis, AR is expressed at similar levels in all SC nuclei, irrespective of the epithelial stage. Note that (i) the left side of each panel corresponds toe staining using the anti-AR antibody and the right side to a DAPI nuclear counterstain and (ii) the histological sections from WT males and from RaraSer−/− mutants were processed in parallel for immunohistochemistry, and identical exposure times were used to acquire the fluorescence pictures. E, elongated spermatids; L, leptotene spermatocytes; LY, Leydig cell; M, spermatocytes in metaphase; P, pachytene spermatocytes; PR, preleptotene spermatocytes; R, round spermatids; S, Sertoli cells; S2, type 2 spermatocytes; Z, zygotene spermatocytes. Roman numerals designate stages of the seminiferous epithelium cycle: II–VI, stage II, III, IV, V or VI; VII–VIII, stage VII or VIII. Bar: 30 μm (A–H). Download figure Download PowerPoint Figure 4.Ablation of RARα in SC abrogates the epithelial stage-dependent variations of Stra6 expression, but does not alter cyclic expression of the germ cell marker Stra8 in adult testes. Immunostaining for STRA6 (A–D) and STRA8 (E–H) in WT and RaraSer−/− testes, as indicated. In WT males, STRA6 protein is present at epithelial stages VII–IX, and its level peaks at stage VIII. In WT testes, immunolabeling for STRA8 is strong in preleptotene spermatocytes (present at stages VII–VIII of the seminiferous epithelium cycle) and weak in leptotene spermatocytes (present at stages IX–X), and STRA8-containing spermatocytes co-distribute with STRA6-containing SC. Note that STRA8 is also expressed in spermatogonia (see Supplementary Figure 7). In RaraSer−/− testes, the epithelial stage-specific expression of STRA6 is lost; in contrast, STRA8 distribution is unaffected. Note that (i) panels A, B and panels E, F correspond to consecutive sections of a WT testis and (ii) that panels C, D and panels G, H correspond to consecutive sections of a RaraSer−/− testis: PR, preleptotene spermatocytes. Roman numerals designate stages of the seminiferous epithelium cycle: II–VI, stage II, III, IV, V or VI; VII–VIII, stage VII or VIII; X–XI, stage X or XI. Asterisks indicate tubule sections containing STRA8-positive leptotene spermatocytes, which are not visible at the illustrated magnification. In panels B, D, F and H, the immunohistochemical signals were converted to a red false color and superimposed with the DAPI nuclear counterstain (blue false color). Bar (in H): 160 μm (A–H). Download figure Download PowerPoint We also analyzed the cyclical activity of SC before the appearance of the seminiferous epithelium cycle. In WT males at P5 and P10, distribution of STRA6 protein, and the cellular retinol-binding protein CRBP1 (Rbp1) and galectin-1 (Lgals1) transcripts varied between seminiferous cord sections (Figure 5A–C and G–I). In contrast, expression of these genes appeared uniform in all seminiferous cords of RaraSer−/− mutants (Figure 5D–F and J–L). Importantly, the total amounts of Stra6, Lgals1 and Rbp1 transcripts measured by quantitative RT–PCR in RaraSer−/− testes at P5 did not significantly differ from those in WT testes (Supplementary Figure 5), indicating that changes in their expression were only qualitative. The GATA1 immunolabeling at P10 varied markedly between individual seminiferous cords in WT testes, whereas it was strong in all SC nuclei in RaraSer−/− testes (compare Figure 5M and N). This abnormal, uniform, expression of GATA1 was also observed in RaraSer−/− testes at P20 (compare Figure 5O and P), indicating that it was not related to the developmental delay of the mutant testis (see above). Altogether, these data demonstrate that the cyclical activity of SC is disrupted in RaraSer−/− mutants, already at the onset of pubertal testis development. Figure 5.Ablation of RARα in SC abrogates the epithelial stage-dependent variations of gene expression in prepubertal testes. (A, D, G, J) Immunostaining for STRA6. (B, C, E, F, H, I, K, L) ISH analyses using Rbp1 (B, E, H, K) and Lgals1 (C, F, I, L) antisense probes; the positive signals for transcripts are violet. (M–P) Immunostaining for GATA1. (A, D, G, J, M–P) The positive signal for STRA6 and GATA1 is red. The DAPI counterstain is also illustrated in panels M–P. Bar (in P): 50 μm (A–F) and 80 μm (G–P). Download figure Download PowerPoint Similar to the adult situation (see above), loss of Stra6 cyclical expression in prepubertal RaraSer−/− testes (Figure 5D and J) did not abolish the cyclical expression of Stra8 (not shown). This finding was quite unexpected, as in prepubertal WT testes we found a strong positive, temporal and spatial, correlation between SC displaying high levels of STRA6- and STRA8-positive spermatogonia and spermatocytes (Supplementary Figure 8). The coordinated expression of Stra6 and Stra8 is therefore uncoupled upon ablation of Rara in SC. Additional ablation of RARβ and RARγ in SC does not increase the severity of the phenotype resulting from RARα ablation A striking variability in the extent of the seminiferous epithelium vacuolation was observed not only in different RaraSer−/− mutants but also within a given mutant in different tubule segments (Figure 1I and Supplementary Figure 3). Although RARα is the only RAR evidenced in SC using immunohistochemistry (Vernet et al, 2006), the variability in seminiferous epithelium degeneration left open the possibility that stochastic variations of RARβ and/or RARγ possibly present in low (i.e., undetectable) amounts could compensate for the RARα loss of function. To investigate this possibility, we generated Rara/b/gSer−/− mice lacking RARα, RARβ and RARγ in SC (see Supplementary information). Testes from 9-week-old Rara/b/gSer−/− mutants (n=3) displayed alterations that were indistinguishable from those found in RaraSer−/− mice (Supplementary Figure 9), indicating that RARα is the sole functional RAR in SC. Ablation of all RXRs in SC does not recapitulate the RaraSer−/− phenotype RXRβ is the predominant RXR in SC (Vernet et al, 2006). Selective ablation of Rxrb in SC (RxrbSer−/− mutation) yielded, at 9 weeks of age, testis defects identical to those generated upon inactivation of RXRβ function in the whole organism (Kastner et al, 1996; Mascrez et al, 2004), namely an accumulation of lipids in SC and a failure of spermiation (our unpublished data). The latter is also generated upon ablation of Rara (see above). On the other hand, testis degeneration (i.e., vacuolation, desquamation of immature round spermatids, increased germ cell apoptosis) and loss of Stra6 cyclical expression, which are hallmarks of age-matched RaraSer−/− mutants, were never observed in RxrbSer−/− mutants (not shown). To exclude the possibility that RXRα and/or RXRγ present at low amounts in SC could compensate for the loss of RXRβ, mice lacking RXRα, RXRβ and RXRγ in SC (i.e., Rxra/bSer−/−/Rxrg-null mutants, see Supplementary information) were analyzed. At 9 weeks of age, these mutants (n=3) recapitulated the defects of RxrbSer−/− mutants (Figure 6D), but did not exhibit degeneration of the seminiferous epithelium (Figure 6B and F) and loss of Stra6 cyclical expression (Figure 6H; not shown). Interestingly, none of the mutations altered the cyclical expression of Stra8 in germ cells (Figure 6I and J). Therefore, in contrast to RARα, RXRs are dispensable for the structural integrity of the seminiferous epithelium and for the cyclical activity of SC. Figure 6.Ablations of all three RARs (i.e., Rara/b/gSer−/− mutants) or of all three RXRs (i.e., Rxra/bSer−/−/Rxrg-null mutants) yield very different abnormalities. Histological sections of testes at 9 weeks of age stained by (A, B) the PAS method, (C, D) oil red O for detection of lipids droplets (red dots), (E, F) the TUNEL method to detect apoptopic cells (green fluorescent signals) and (G–J) ISH using antisense probes to (G, H) Stra6 and (I, J) Stra8 (purple signals). (A, C, E, G, I) The Rara/b/gSer−/− mutant testis displays (i) large, lipid-free, vacuoles (VA in panel A), (ii) numerous apoptotic round germ cells (some indicated by yellow arrowheads in panel E) and elongated spermatids (indicated by white arrowheads in panel E) in numerous seminiferous tubules sections (T) and (iii) low and uniform expression of Stra6 in all seminiferous tubules that does not follow the cyclical expression of Stra8 shown on an consecutive section. (B, D, F, H, J) In cont
Referência(s)