LRH-1 (liver receptor homologue 1/NR5A2) is an orphan member of the nuclear receptor family that is highly expressed in liver, pancreas, colon, and intestine where it plays an important role in the control of bile acid synthesis and homeostasis (Lu et al., 2000, Repa and Mangelsdorf, 2000). LRH-1 regulates the expression of rate-limiting enzymes in bile acid synthesis, including cholesterol-7α-hydroxylase and sterol 12α-hydroxylase (Nitta et al., 1999, Del Castillo-Olivares and Gil, 2000; products of the CYP7A1 and CYP8B1 genes, respectively). LRH-1 is closely related in structure to steroidogenic factor 1 (SF-1/NR5A1) another orphan nuclear receptor. Both are members of the Fushi Tarazu Factor 1 gene family and share a high degree of sequence similarity in their DNA-binding domains. Thus, both nuclear receptors can bind to the same hexameric response elements in DNA (Galarneau et al., 1996) and theoretically are capable of activating the same genes.
LRH-1 also is expressed in the gonads. In ovaries, high levels of expression have been reported in the adult mouse (Repa and Mangelsdorf, 2000), horse (Boerboom et al., 2000), and humans (Sirianni et al., 2002). We and others reported that, in rodent ovary, LRH-1 is localized specifically to granulosa cells of the follicle and to cells of the corpus luteum (Liu et al., 2003, Falender et al., 2003, Hinshelwood et al., 2003, Mendelson et al., 2005), suggesting that it could potentially play a role in the regulation of ovarian steroidogenesis. We recently reported that spatial and temporal expression of LRH-1 in the cycling and pregnant mouse ovary was correlated with that of aromatase P450 (p450arom, product of the cyp19 gene), whereas, SF-1 expression was essentially undetectable in the pregnant mouse ovary and more closely associated with expression of 17α-hydroxylase/17,20 lyase (p45017α) during the estrous cycle (Mendelson et al., 2005). In contrast to expression in the ovary, levels of expression of LRH-1 were much lower in adult rodent testes (Repa and Mangelsdorf, 2000, Pezzi et al., 2004). In contrast to our knowledge of LRH-1 expression and function in adult gonads, little is known regarding LRH-1 expression during gonadal development.
Development of the gonads requires coordinated expression of several transcription factors early in formation of the urogenital ridge and later in the presumptive gonad. In mice, primordial gonads emerge ventromedial to the mesonephros. These gonadal primordia are bipotential and express several transcription factors, including SF-1 (Brennan and Capel, 2004). Besides its important role in gonadal development (Luo et al., 1994), SF-1 regulates steroidogenic enzymes responsible for formation of testosterone within the developing testis (Habert et al., 2001); testosterone, in turn, plays a critical role in differentiation of the internal and external genitalia. Moreover, in the adult, SF-1 performs an essential function in regulating steroidogenesis in the testis, ovary, and adrenal (Ikeda et al., 1993, Morohashi et al., 1995). In these respects, SF-1 is thought to be a competence factor necessary for steroidogenic function, both during development and in the adult.
Mice with a targeted deletion in the sf1 gene are born without gonads, adrenals, pituitary gonadotrophs, and ventromedial nucleus in the hypothalamus (Luo et al., 1994). On the other hand, mice homozygous for a germline mutation of the lrh1 gene die between embryonic day (E) 6.5 and E7.5 in part because LRH-1 is necessary for formation of endoderm (Pare et al., 2004, Gu et al., 2005). Thus, a role for LRH-1 in gonadal development remains to be established. To gain insight into the role of LRH-1 vs. SF-1 in embryonic and postnatal gonadal development and maturation, in the present study, we (1) analyzed the pattern of expression of LRH-1 vs. SF-1 throughout embryonic and postnatal gonadal development, and (2) compared cell-specific patterns of expression of LRH-1 and SF-1 with those of putative target genes of LRH-1 and/or SF-1, steroidogenic enzymes P450arom and P45017α.
LRH-1 Is Expressed in the Indifferent Gonad in a Temporal and Spatial Pattern That Differs From SF-1
To determine the expression pattern of LRH-1 during gonadal development and to compare it with that of SF-1, in situ hybridization for both transcripts was performed on serial sections of mouse embryos at various stages of development. At E10.5, just before the time when the gonadal primordium can be recognized, signal for LRH-1 could not be distinguished from background in urogenital ridges (Fig. 1A,B). By E11.5, when the bipotential gonad can be differentiated from adrenal primordium, signal for LRH-1 was evident in gonad, but not the adrenal primordium (Fig. 1A). At higher magnification, transcripts for LRH-1 were evident in a subset of the discernible germ cells (large round cells with large nuclei and relatively little cytoplasm) and in somatic cells surrounding these positive germ cells (Fig. 1B). Signal was sparse in epithelium of the genital ridge.
In contrast to LRH-1, signal for SF-1 was evident throughout the urogenital ridge at E9.5 (data not shown), E10.5, and E11.5 (Fig. 1A), as has been reported previously (Ikeda et al., 1994). Signal for SF-1 was detected in somatic cells throughout the ridge at E10.5. Unexpectedly, however, at E11.5, SF-1 signal was evident also in germ cells.
LRH-1 and SF-1 Are Differentially Expressed in Developing Male and Female Gonads
By E13.5 testis and ovary are clearly anatomically distinct. The testicular cords contain cells destined to become Sertoli cells and primordial germ cells that will become spermatogonia. The interstitium is composed of myoid and Leydig cell precursors (Habert et al., 2001). At E13.5, LRH-1 transcripts were present only in testicular cords (Fig. 1A,B). At high magnification, signal was primarily evident in germ cells and in pre-Sertoli cells (triangular-shaped cells; Fig. 1B). On E15.5, there was an overall decline in the expression of LRH-1 in testes. This decline was not secondary to poor fixation, as signal in the intestine remained robust (Fig. 1A). In addition, this decrease in testicular expression was observed in more than one series of experiments.
At E13.5, signal for SF-1 was detected in both testicular cords and interstitium; however, the abundance of signal was much greater in the interstitium than in the cords (Fig. 1A,B). In the testicular cords, signal was observed only in the pre-Sertoli cells (Fig. 1B). At E15.5, signal for SF-1 was maintained, primarily in the interstitium (Fig. 1A). In the adrenal, extremely low levels of SF-1 expression were detected, beginning at E10.5 and remained evident at E13.5 (Fig. 1A) and E15.5 (data not shown).
On E13.5, LRH-1 signal was evident throughout ovary, but absent in overlying epithelium (Fig. 1A). At higher magnification, signal was localized to germ cells within germline cysts (Pepling and Spradling, 1998, 2001) and to somatic cells in contact with germ cells (Fig. 1B). As observed in the developing testis, signal for LRH-1 declined dramatically in the ovary to background levels at E15.5 (Fig. 1A). This decrease contrasts to expression observed in pancreas, which was intense both at E13.5 (Fig. 1A) and E15.5 (data not shown). The finding of high pancreatic expression is in agreement with previously published reports (Galarneau et al., 1996).
At E13.5, SF-1 was expressed at relatively high levels throughout the ovary and was more robust than LRH-1 (Fig. 1A). SF-1 expression was evident both in somatic cells and in epithelium surrounding the ovary (Fig. 1B). By E15.5, SF-1 signal was decreased compared with levels seen at E13.5 (Fig. 1A). This decline in SF-1 levels also has been reported in the ovaries of mice (Ikeda et al., 1994) and rats (Shen et al., 1994) during embryogenesis.
LRH-1 Expression in the Testes Remains Low During Postnatal Development
Although LRH-1 transcripts were low in postnatal day (P) 2 testes (Fig. 2), expression in germ cells within seminiferous tubules was clearly evident at high magnification (Figs. 2, 3A). By P8, when proliferation of spermatogonia and meiosis has begun (McCarrey and Abbott, 1979), signal for LRH-1 decreased to barely detectable levels (Fig. 2). At P24 and in adult testes, LRH-1 mRNA was at background levels.
In contrast, signal for SF-1 was clearly evident and localized to the Sertoli and interstitial cells on P2 (Figs. 2, 3B). By P8, signal for SF-1 was observed both in the Sertoli cells and the interstitium, with the most abundant signal in the interstitium (Fig. 2). In the P24 and adult testes, SF-1 was primarily expressed in the Leydig cells (Fig. 2). In addition, there was detectable signal in the Sertoli cells, although the signal was greatly decreased compared with earlier postnatal time points (data not shown).
P45017α mRNA Expression Correlates With That of SF-1 in the Postnatal Testes
In the testes at P2 and P8, intense signal for P45017α was restricted to the interstitium (Fig. 2), particularly in a population of peritubular cells that appear to be precursors of adult Leydig cells (Habert et al., 2001). At P24 and in adult testes, P45017α signal was still restricted to Leydig cells (Fig. 2). This finding is quite similar to the expression pattern of SF-1 throughout postnatal development of the testes (Fig. 2). Quite unexpectedly, transcripts for P45017α were highly expressed in the epithelium of the epididymis, but only at P24. It is not clear what transcription factors may be responsible for regulation of expression of P45017α in the epididymis at P24, as we could not detect SF-1 or LRH-1 in this tissue. Moreover, it is unclear what role P45017α expression may play in the epididymis at this time point.
P450arom and LRH-1 Transcripts Are Not Detectable in the Postnatal Testes
Transcripts for P450arom were not detectable in the P2 or P8 testes (Fig. 2) and have been reported to be low at all stages of male development (Abney, 1999). This is not surprising, because estrogen formed by the action of the aromatase enzyme complex is thought to be inhibitory to Leydig cell development (Abney, 1999). On the other hand, on P24 and in adult testes, individual tubules were found to contain signal for P450arom (Fig. 2). P450arom is known to be expressed both in Sertoli cells before puberty, and subsequently at low levels in Leydig cells (Abney, 1999; Carreau, 2002). P450arom expression also has been reported in male germ cells, including pachytene spermatocytes, round spermatids to elongated spermatids, and in spermatozoa (Carreau, 2001).
Expression Pattern of LRH-1 in the Early Postnatal Ovary Correlates With the Initiation of Folliculogenesis
The perinatal ovary contains naked oocytes and primordial follicles composed of an oocyte surrounded by one layer of flattened granulosa cells (Pedersen and Peters, 1968). The transition of primordial to primary follicle is marked by the proliferation of granulosa cells (2–5 to ≤ 20 cells) and a change in their shape, an increase in oocyte size and formation of the zona pellucida (Lintern-Moore and Moore, 1979). Two days after birth, oocytes within these primordial follicles are arrested in prophase I of the first meiotic division, which is not completed until ovulation. In the mouse, folliculogenesis begins just before birth in the medullary portion of the ovary. Transcription factors such as Factor in Germline α (Soyal et al., 2000) and Wnt4 (Vainio et al., 1999) have been postulated to play a role in formation of primordial follicles; however, the mechanism by which some primordial follicles grow and others stay quiescent is still unknown. Thus, it is of great interest that the induction of LRH-1 expression appears to be correlated with the initiation of follicle growth and development.
In the perinatal ovary, growing follicles are located in the medullary region, whereas most of the quiescent, nongrowing primordial follicles and naked oocytes are located in the cortex (Peters et al., 1975). Additionally, there are still some germline cysts present in the cortex. Overall, the majority of signal for LRH-1 was evident in the medullary portion (Fig. 4). Both in the cortex and medulla, naked oocytes were found that manifested no signal for LRH-1 (Fig. 3C, follicle ). In primordial follicles, both the oocyte and surrounding granulosa cells were positive for LRH-1 (Fig. 3C, follicle ). A small number of single-layer primary follicles observed in the medulla also displayed signal for LRH-1 primarily in the granulosa cells and to a lesser extent in the oocyte (Fig. 3C, follicle ). Thus, during the initiation of folliculogenesis, the pattern of LRH-1 expression appears to change in follicles as they begin to grow (Figs. 3C, 5).
By P8, many primordial follicles have developed into primary and multilayered primary follicles. Single-layer primary follicles exhibited signal for LRH-1 both in oocytes and in granulosa cells. However, with the transition to multilayer primary and larger follicles, signal was detected only in granulosa cells (Fig. 4). In contrast to the P2 ovary, there were several primordial and primary follicles that had low levels of signal for LRH-1; however, these follicles appeared to be atretic. In the P24 and adult ovary, signal for LRH-1 was present in granulosa cells of follicles from primary to antral stage (Fig. 4). Follicles that had morphological signs of atresia had decreased signal for LRH-1. No signal was detected in oviduct (Fig. 4).
In contrast to LRH-1, SF-1 was evenly distributed throughout the ovary at P2 (Fig. 4). At higher magnification, signal was evident in granulosa cells of follicles and in mesenchyme within the ovary (Fig. 3D). Granulosa cells of both primordial and primary follicles demonstrated signal for SF-1; however, it was not correlated with stage of follicular development (Fig. 3D). At P8, there was a high level of SF-1 signal throughout the medullary portion of the ovary, where the growing follicles are located (Fig. 4). This signal was present in granulosa cells and developing thecal layer. Moreover, the cortical zone containing primordial, nongrowing follicles (Peters et al., 1975) lacked signal for SF-1. Transcripts for SF-1 were present in both cortex and medulla of the ovary at P24 and in adult, in which the strongest signal was in theca interna and interstitium (Fig. 4).
Cell-Specific Expression of P45017α and P450arom Correlates With Changes in the Expression of LRH-1 and SF-1
On postnatal day 2, there was no detectable expression of either P45017α or P450arom in the ovary (Fig. 4). This finding was to be expected, as the majority of primordial follicles are quiescent. By P8, there was abundant signal for P45017α in the interstitium around the follicles (Fig. 4). This finding coincides with formation of multilayered primary follicles and with organization of the mesenchyme surrounding these follicles into theca interna. Just before puberty, at P24, and in adult ovary, P45017α mRNA was abundant both in the interstitial region and in theca interna (Fig. 4). This finding correlated with high expression of SF-1 in the interstitium and theca interna. For this reason, it appears that SF-1 serves as a competence factor for P45017α, as has been suggested previously (Zhang and Mellon, 1997, Mellon et al., 1998).
P450arom expression on P8 was readily detectable in granulosa cells of multilayered primary follicles (Fig. 4). By P24, abundant signal for P450arom was present in all antral follicles (Fig. 4). However, in the adult, only a subset of antral follicles contained signal for P450arom (Fig. 4). At all stages of development where P450arom was expressed in granulosa cells, signal for both LRH-1 and SF-1 also was present. Of interest, both LRH-1 and SF-1 were present in the cumulus cells that surround the oocyte, but P450arom was not, indicating that either additional factors must be necessary for P450arom expression in these cells or that a factor(s) possibly secreted by the oocyte may inhibit P450arom expression.
Onset of Expression of LRH-1 in Primordial Germ Cells and Somatic Cells in the Urogenital Ridge Is Correlated With Distinct Developmental Events
In the present study, transcripts for LRH-1 were first detected at E11.5 in primordial germ cells (PGCs) and surrounding somatic cells of the urogenital ridge. PGCs double to quadruple in number between E9.5 and E11.5 (Buehr, 1997). Between E10.5 and E11.5, these cells lose their ability to migrate and, thus, colonize the urogenital ridge (Molyneaux et al., 2001, 2004). The coordinate induction of LRH-1 expression in PGCs with their arrival at the urogenital ridge raises the question of whether LRH-1 expression is induced in PGCs upon colonization of the ridge. Additionally, it is of interest that SF-1 is transiently expressed in PGCs at E11.5. Thus, it will be of importance to determine the role of LRH-1 and SF-1 in PGC function.
At E11.5, signal for LRH-1 was observed in both PGCs and in somatic cells of the urogenital ridge. However, unlike the expression pattern for SF-1, signal for LRH-1 was sparse in coelomic epithelium. During testis development, SF-1–expressing cells of coelomic epithelium rapidly proliferate and then migrate into the urogenital ridge where they give rise to both Sertoli and interstitial cells (Brennan and Capel, 2004). Although it appears that LRH-1 is transiently coexpressed with SF-1 in cells that are destined to become Sertoli cells, LRH-1 is never observed in Leydig cells. The role that LRH-1 may play in Sertoli cell differentiation and development is a question for future study.
LRH-1 May Play a Role in Proliferation of PGCs in the Developing Gonad
LRH-1 expression remained elevated upon differentiation of both testes and ovaries. In the testes at E13.5, signal for LRH-1 was localized within the testicular cords to both somatic cells and PGCs. This finding is in contrast to SF-1, which was expressed highly in the interstitium as well as in the pre-Sertoli cells in the testicular cords. LRH-1 expression in the testes declined after E13.5. This timing corresponds to the time when PGCs have stopped dividing and are arrested in the G1/G0 phase of the cell cycle (McLaren, 2000). Recently, LRH-1 was shown to have potent effects on intestinal cell proliferation through the simultaneous induction of cyclin D1 and E1 in intestinal crypt cells, suggesting a role of LRH-1 in their renewal (Botrugno et al., 2004). The expression pattern of LRH-1 in germ cells of the developing gonad observed in the present study suggests that it may play a role in germ cell proliferation.
In the developing ovary, LRH-1 expression appears to be spatially and temporally associated with mitosis of the PGCs. Signal for LRH-1, which was relatively high on E13.5 in PGCs and in surrounding somatic cells of germline cysts, declined to background levels on E15.5. PGCs undergo mitosis between E10.5 to ∼E14.5, with the highest rate of mitosis occurring just before the beginning of meiosis on ∼E13.5 (Byskov, 1986). The PGCs divide synchronously from a progenitor cell; by incomplete cytokinesis, they remain interconnected to form germline cysts (Pepling and Spradling, 1998). Once mitosis ends, PGCs in these cysts begin the process of meiosis and eventually arrest in the diplotene stage of meiosis I during the perinatal period. Some of the somatic cells begin to associate closely with PGCs in these cysts beginning on E14.5 (Byskov, 1986); however, it is not until E18.5, when these germline cysts begin to undergo apoptosis, that primordial follicles begin to form (Pepling and Spradling, 2001). Interestingly, LRH-1 expression was found to be reinitiated in the postnatal ovary on day 2. Thus, LRH-1 expression appears to be spatially and temporally associated in developing ovary with mitosis of the PGCs and then, after birth, with formation of the primordial follicle. Moreover, as observed in the E11.5 urogenital ridge, signal for LRH-1 was primarily localized to those somatic cells in contact with PGCs. Thus, germ cell interaction with somatic cells may play a role in LRH-1 expression in early ovarian development.
LRH-1 May Play a Role in the Initiation of Folliculogenesis in the Postnatal Ovary
Increased expression of LRH-1 in the ovary on postnatal day 2 appears to be correlated with initiation of folliculogenesis (Fig. 5). LRH-1 transcripts were localized to primordial and to primary follicles containing a single layer of granulosa cells within the medullary region of the ovary, which contains the more highly developed follicles (Fig. 3C). By contrast, LRH-1 signal was not detected in naked oocytes, lacking granulosa cells. These oocytes are located primarily in the cortical region of the ovary (Figs. 3C, 5). Primordial follicles are composed of a few flattened granulosa cells that surround the oocyte. These granulosa cells are thought to arise from surrounding mesenchyme (Lintern-Moore and Moore, 1979). Although there are several factors that appear to be important for formation of primordial follicles and/or its transition to primary follicles, these factors are either expressed in oocytes or granulosa cells but not in both cell types (Richards, 2001; Epifano and Dean, 2002). Thus, it is of great interest that LRH-1 is expressed in both oocytes and granulosa cells of primordial and single-layer primary follicles, because communication between oocyte and granulosa cells is known to occur (Albertini and Barrett, 2003). In light of the potential role of LRH-1 in regulation of proliferation of intestinal cells (Botrugno et al., 2004), one might hypothesize that LRH-1 produced by oocytes and/or by granulosa cells could regulate granulosa cell proliferation during folliculogenesis.
Whereas, LRH-1 expression is present in oocytes of primordial and single-layer primary follicles, it is absent in oocytes of larger follicles. Germ cell nuclear factor 1 (GCNF-1), another orphan member of the nuclear receptor family, also is expressed in oocytes of developing primary and more mature follicles but not in primordial follicles (Katz et al., 1997). Thus, LRH-1 expression begins to decline in primary follicles when GCNF-1 expression is turned on. A similar pattern of reciprocal expression between LRH-1 and GCNF-1 was reported recently in the developing embryo and during embryonic stem cell differentiation, where both factors maintain Oct-4 expression (Gu et al., 2005). The mechanisms for the reciprocal expression of these nuclear receptors in oocytes and consequences on target gene expression are questions for future study.
LRH-1 Plays a Potential Role in Regulation of Steroidogenesis in Postnatal Gonad
Upon testicular differentiation (E13.5), LRH-1 was only transiently expressed in germ and Sertoli cells; signal for LRH-1 was never observed in Leydig cells. Conversely, SF-1 expression was maintained in both Sertoli cells of the tubules and Leydig cells of the interstitium throughout the prenatal and postnatal period. P45017α expression was observed in the interstitium in fetal testis and in adult Leydig cells. Thus, the expression pattern of SF-1 appeared to mirror that of P45017α throughout the prenatal and postnatal period. For this reason, it appears that SF-1 serves as a primary competence factor for P45017α expression in testis, as has been suggested previously (Givens et al., 1994).
After differentiation of the ovary at E13.5, transcripts for both LRH-1 and SF-1 declined. LRH-1 expression was again increased in the ovary after birth on P2. SF-1 expression was reported to increase just before birth on E18.5 in the mouse (Ikeda et al., 1994) and just around the time of birth in the rat (Shen et al., 1994). In the mouse ovary, unlike the testis, steroidogenesis does not occur until after birth. Using reverse transcriptase-polymerase chain reaction (PCR), levels of P450arom were detected on P1, with the biggest increase occurring from P1 to P5 (Gray et al., 1995). In the present study, we first observed P450arom and P45017α transcripts in the ovary on P8 in granulosa and theca layers of multilayer primary follicles, respectively. LRH-1 expression was restricted to granulosa cells of these follicles, whereas SF-1 was present in both granulosa and theca cell layers. Thus, in granulosa cells, both LRH-1 and SF-1 are present and correlated with P450arom expression, whereas in the theca, SF-1 expression is correlated with that of P45017α.
Of interest, the pattern of P450arom expression varied with stage of follicular development and postnatal age. Whereas, the multilayer primary follicles expressed P450arom mRNA at P8, these multilayer primary follicles lacked signal for P450arom in ovaries from P24 and adult mice. The cause for this difference is unknown; however, most of the primary follicles in the P8 ovary manifested no visible signs of atresia and may have been undergoing a period of rapid growth. Conversely, in ovaries of older mice, many of the primary follicles may have been undergoing atresia. Similarly, in ovaries obtained from mice before puberty at P24, all of the large follicles manifested signal for P450arom, whereas in ovaries from adult mice, only a subset of preovulatory follicles demonstrated signal for P450arom. The absence of P450arom expression in some of the large follicles in ovaries from adult mice also correlated with signs of follicular atresia. Thus, in follicles from all stages of postnatal development, P450arom expression appears to be correlated both with follicle health and with expression of LRH-1 and SF-1.
In summary, two closely related orphan members of the nuclear receptor family, LRH-1 and SF-1, were found to be differentially expressed in the indifferent gonad and in developing ovaries and testes. The absence of gonadal development in mice with a targeted deletion of the sf1 gene suggests an important role of SF-1 in this process. On the other hand, the role of LRH-1 in gonadal development is uncertain, because lrh1 null embryos die at E6.5–E7.5 (Pare et al., 2004; Gu et al., 2005). The findings presented herein suggest that LRH-1 may play a role in primordial germ cell proliferation within the indifferent gonad, in folliculogenesis in the early postnatal ovary and in the regulation of estrogen biosynthesis. Additionally, LRH-1 may serve an important function in developing Sertoli and granulosa cells. Understanding of the downstream targets of LRH-1, as well as the factors that regulate LRH-1 expression, may provide greater insight into its role in gonadal development and function.
MATERIALS AND METHODS
Preparation of Mouse Embryos and Tissues
Mice were housed under a 12-hr light cycle (lights on, 6:00 AM to 6:00 PM) at 22°C. They were treated in accordance with the guidelines set forth by the Animal Welfare Information Center, and protocols were approved by the Institutional Animal Care and Use Committee at University of Texas Southwestern Medical Center at Dallas. Embryos were harvested from ICR mice (Harlan, Indianapolis, IN) as described previously (Hinshelwood et al., 2003). Briefly, timed-pregnant ICR mice were anesthetized and then killed by cervical dislocation. Embryos were dissected from the uterus, rinsed in cold diethylpyrocarbonate (DEPC) -saline, and immediately fixed in chilled 4% formaldehyde (PFA, freshly prepared from paraformaldehyde) in DEPC–phosphate buffered saline (PBS), pH 7.4. The exact age of these embryos was later determined by enumerating somites and evaluating anatomic features of limbs.
Postnatal ovaries and testes were collected from ICR mice at 2, 8, and 24 days of age and from adults followed by fixation as described previously (Hinshelwood et al., 2003). Briefly, animals were anesthetized, and transcardial perfusion was performed with cold DEPC–heparinized saline followed by chilled 4% PFA in DEPC–PBS, pH 7.4. Tissues (including embryos) were then dissected and fixed for 16 hr with rocking in 4% PFA at 4°C before transfer to DEPC–PBS. Tissues were then dehydrated, paraffin embedded, and sectioned at 5 μm onto microscope slides treated with Vectabond (Vector Laboratories, Burlingame, CA). Slides were stored desiccated at 4°C until use for in situ hybridization.
At the time of sectioning, to confirm sexual genotype, a modification of a standard protocol was used for DNA extraction (http://cc.ucsf.edu/people/waldman/protocols/paraffin.html). Briefly, five 10-μm paraffin sections were collected in an Eppendorf tube and deparaffinized with l ml of xylene. After centrifugation and decanting of supernatant, a second paraffin extraction was performed followed by two successive ethanol washes and centrifugations. The resulting pellet was air-dried before proteinase K digestion. A 0.5-ml volume of “digestion buffer” (Alcorn et al., 1993) was added to the sample, and the sample was incubated overnight at 55°C with shaking. After protein digestion, equal volumes of phenol–chloroform–isoamyl alcohol were added, extracted, and centrifuged. The aqueous supernatant was removed to a new tube, and this step was repeated if needed. Glycogen was added to the aqueous phase together with 2–2.5× volume of 100% ethanol, and the sample was precipitated overnight. After centrifugation, the resulting pellet was washed with 70% ethanol and air-dried before resuspension in 20 μl of water.
Genotyping was accomplished using 1–3 μl of isolated DNA (depending upon embryo age) and primers specific for the Y chromosome-specific genes, zinc finger protein 1, and zinc finger protein 2 (zf1;22767 and zf2;22768). The primers corresponded to regions that recognized both genes (a gift from Dr. Pancharatnam Jeyasuria): forward primer, 5′-CCT ATT GCA TGG ACA GCA GCT TAT G-3′ and reverse primer, 5′-GAC TAG ACA TGT CTT AAC ATC TGT CC-3′. The PCR conditions were as follows: 94°C for 4 min (1 cycle); 94°C for 40 sec, 50°C for 40 sec, 68°C for 2 min (40 cycles); 68° C for 10 min (1 cycle). A portion of the PCR amplicon was run on a 1% gel (Agarose-1000; Invitrogen Co., Carlsbad, CA) and visualized using ethidium bromide and a ultraviolet light.
Preparation of Probes for In Situ Hybridization
Templates for mouse SF-1 and LRH-1 probes were subcloned into BlueScript KS+ (Stratagene Co., La Jolla, CA), as described previously (Hinshelwood et al., 2003). Briefly, the SF-1 probe corresponded to a 200-bp fragment of the 3′-untranslated region of mouse SF-1, a gift from Keith L. Parker (UT Southwestern; Ikeda et al., 1993), whereas the mouse LRH-1 probe corresponded to a 502-bp fragment containing the ligand binding domain. Mouse P45017α and P450arom probes (Burns et al., 2001) were a gift from Martin M. Matzuk (Baylor College of Medicine, Houston, TX). 35S-labeled cRNA probes for in situ hybridization were synthesized using a T3/T7 MaxiScript kit (Ambion, Inc., Austin, TX) and [35S]-UTP (1,250 Ci/mmol; Kamat et al., 1999). These were hybridized to tissue sections as described previously (Shelton et al., 2000). After hybridization, slides were coated with K.5 nuclear emulsion (Ilford, UK), exposed at 4°C for 3–6 weeks, developed, counterstained with hematoxylin, and photographed by using bright- and darkfield optics.
The authors thank Dr. Pancharatnam Jeyasuria for assistance with genotyping and Chris Pomajzl, Jeffrey Stark, and David Sutcliffe for their assistance with histology. M.M.H. was funded in part by an Obstetrics/Gynecology Basic Research Fund.