Sex-specific differences in mouse DMRT1 expression are both cell type- and stage-dependent during gonad development

Abstract

Immunohistochemistry was used to examine GCNA1, a germ cell specific protein, together with DMRT1, a transcription factor implicated in Sertoli cell and germ cell function, in order to resolve DMRT1’s cellular profile during pre- and postnatal gonad development in the mouse. In the indifferent gonad (10.5–11.5 dpc), DMRT1 localized to somatic cells and GCNA1⁺ germ cells and was indistinguishable in males and females. By 12.5 dpc, a clear sexual preference for DMRT1 in male somatic cells was observed, with male DMRT1 localized to testicular cords and more abundant in Sertoli cells than germ cells and female DMRT1 diffusely labeled and markedly lower in somatic cells than germ cells. A male somatic preference continued throughout development, with DMRT1 evident in Sertoli cells at all ages examined and absent in ovarian somatic cells from 13.5 dpc onward. In contrast, expression in primordial germ cells was not sexually distinct and both sexes showed DMRT1 increasing through 13.5 dpc and absent by 15.5 dpc. Notably, sexual differences in germ cell DMRT1 were revealed after birth, when it was detected only in spermatogonia of the testis. Co-localization of DMRT1 with proliferation markers KI67 and PCNA and stem cell markers OCT4 and NGN3 indicated that, in postnatal testes, DMRT1 was present in both stem and proliferating spermatogonia. Together, the findings implicate opposite functions for DMRT1 in somatic and germ cells of the testis. In Sertoli cells, DMRT1 expression correlated with differentiation, while in germ cells it suggested a role in expansion and maintenance of undifferentiated spermatogonia.

INTRODUCTION

DMRT1 (Doublesex and Mab-3 related transcription factor-1) is a putative transcription factor required for testis differentiation and is the presumed vertebrate ortholog of Doublesex and Mab-3, two evolutionarily conserved proteins that direct similar aspects of sexual differentiation in D. melanogaster and C. elegans, respectively [1–3]. These transcription factors share a common DNA binding motif, the DM domain, which is both structurally and functionally conserved, indicating the existence of a common ancestral gene and a conserved molecular pathway that directs aspects of sexual differentiation [4]. In addition to its featured DM domain and evolutionary link to sexual differentiation, DMRT1’s characterized expression profiles further implicated it in the regulation of testis development [2;5]. Thus, in all vertebrates examined to date, Dmrt1 (or its Medaka ortholog Dmy) was restricted to the gonads, suggesting its conserved activities are affiliated with gonad differentiation and function [5–13]. Notably, during gonad development, Dmrt1’s expression profiles exhibited a dynamic, sexually-dimorphic pattern that signified a role in testis differentiation. However, the developmental stage of the observed dimorphic pattern depended upon the species, suggesting phylogenetic differences in Dmrt1 function. Thus, in some vertebrates, Dmrt1’s expression profile suggested it acts during sex determination, while in others, it suggested a dominating influence after sex determination. For example, in some lower vertebrates (e.g., fish, frog, alligator, turtle, and chicken), expression of Dmrt1 (or Dmy) prior to and during sex determination was higher in males than females, while in mammals, dimorphic expression appeared after sex determination [14–18].

In mammals, Dmrt1 expression has been studied most extensively in mice and mainly by examining mRNA [5;9;19]. At early stages of mouse development, Dmrt1 was observed exclusively in the genital ridge of XX and XY embryos, with similar mRNA levels observed throughout the sexually indifferent stage (i.e. up to 11.5 days post coitum, dpc) [5]. By 12.5 dpc and 13.5 dpc, a point when male and female gonads are morphologically distinct, a clear difference in the level of Dmrt1 message was noted; in the testis, expression was restricted to the testicular cords; whereas in the ovary, it was more diffuse. After 13.5 dpc, Dmrt1 levels intensified in the testis and declined in the ovary [9;10;19;20]. Thus, while the dimorphic expression pattern of Dmrt1 suggested a predominant role in male sexual development, its occurrence after the gonads begin their sex-specific programs argues against a role in the initial determination of gonadal sex. Notably, Dmrt1’s implicated function subsequent to sex determination was substantiated by observations in Dmrt1-deficient mice, which displayed testicular defects only in postnatal mice [18].

While these studies initially identified Dmrt1’s sexually dimorphic profile, they revealed only mRNA levels and provided limited information on DMRT1 production in specific cells of the developing gonads. Subsequent immunohistochemical analyses confirmed DMRT1 protein in 18.5 dpc testes, where it was observed only in Sertoli cell nuclei, with no detectable protein in germ cells [18]. After birth, limited analysis of DMRT1 expression in the testis showed it within Sertoli cells and germ cells. What remains uncertain, however, is the cellular profile of DMRT1 during early gonad development, particularly during sex determination, as previous studies did not elucidate distinct cellular contributions to the embryonic expression profile. Furthermore, there is little available information on DMRT1’s cellular profile in the postnatal ovary and testis, in particular the expressing germ cells of the testis. Thus, to extend our understanding of DMRT1, immunohistochemistry was used to evaluate its cellular profile, by comparing its expression to Germ Cell Nuclear Antigen (GCNA1), a protein that specifically marks germ cells, KI67 and Proliferating Cell Nuclear Antigen (PCNA), markers of cell proliferation, and OCTAMER-4 (OCT4, POU5F1) and NEUROGENIN-3 (NGN3), markers of spermatogonial stem cells [21–27]. The findings showed that sexually dimorphic expression of DMRT1 in supporting cells initiated shortly after sex determination and continued thereafter. In contrast, sexual differences in germ cell expression were not evident until after birth, when DMRT1 was observed only in type A spermatogonia of the testis [23;24].

MATERIALS AND METHODS

DMRT1 antibody production

DMRT1 rabbit polyclonal antibody was prepared by Covance Research Products Inc. (Denver, PA), using a 15 amino acid peptide antigen, corresponding to the carboxy-terminal region of rat DMRT1 (CSYAVNQVLEEDEDE-COOH), conjugated to the carrier protein Keyhole Limpet Hemocyanin. Immune serum from two separate preparations, KS213 and KS214, and the corresponding preimmune serums were supplied and used for analysis. Both KS213 and KS214 antibody preparations behaved identically in Western blot analysis and immunohistochemistry. Pre-absorbed DMRT1 antibody was prepared by incubating antiserum with the peptide antigen (100 μg peptide/ml serum) at 4°C overnight.

Animals

Mouse genetic backgrounds were either C57BL/6 (B6), B6SJLF1 hybrid or Swiss Webster. Embryos were staged using the morning of an observed vaginal plug as 0.5 dpc. For mice 15.5 dpc and older, sex was determined by gonad morphology. For embryos between 10.5–15.5 dpc, sex was determined by the presence or absence of Sry. For these embryos, DNA was isolated from the embryonic yolk sac and PCR-amplified using Sry-specific primers (Sry-forward 5′-AAGCGCCCCATGAATGCAT-3′, Sry-reverse 5′-CGATGAGGCTGATATTTATA-3′). Male embryos were identified by the presence of a 216 bp product. Mice were maintained on a 12 hour light-dark cycle and given food and water ad libitum. All animals were cared for in accordance with National Institutes of Health guidelines and experimental procedures were approved by the Laboratory Animal Research Committee at the University of Kansas Medical Center.

Immunohistochemistry

Embryos were isolated and placed in phosphate buffered saline (PBS; 0.137 M NaCl, 2.7 mM KCl, 8.0 mM Na₂HPO₄, 2.0 mM KH₂PO₄, pH 7.4) until fixation. Whole embryos or isolated postnatal gonads were fixed in 4% paraformaldehyde/PBS overnight at 4°C and embedded in paraffin. 5 μm sections were cleared in xylene, rehydrated through graded alcohol solutions, and heated in 10 mM sodium citrate buffer (pH 6.0) with microwaves (1 min at full power, followed by 9 min at medium power) for antigen retrieval. Sections were incubated in Peroxo-Block (Zymed Laboratories Inc. San Francisco, CA) at room temperature for 45 sec and blocked for 60 min at room temperature in 10% normal goat serum blocking solution (Zymed Laboratories Inc.). Primary antibody incubations were carried out overnight at 4°C in blocking solution (1:4000 dilution of rabbit anti-DMRT1 antibody KS213). For enzymatic staining, sections were incubated for 10 minutes at room temperature with a biotinylated goat anti-rabbit secondary antibody (Zymed Laboratories Inc.). Following three 5 minute washes in PBST (PBS buffer with 0.1% Triton X-100), sections were incubated at room temperature for 10 minutes with histostain-plus streptavidin peroxidase (Zymed Laboratories Inc.) and then washed 3 times, 5 minutes each, in PBST. AEC substrate solution (Zymed Laboratories Inc.) was applied to the sections for 5–10 minutes and the reaction stopped by rinsing in PBS. Sections were then lightly counterstained with hematoxylin and mounted on glass slides with GVA-mounting solution (Zymed Laboratories Inc.).

For immunofluorescence co-staining, primary antibody incubations were carried out overnight at 4°C in blocking solution using the following dilutions: 1:4000 of rabbit anti-DMRT1 antibody; 1:4 of rat anti-GCNA1 serum 10D9G11 (kindly provided by Dr. G. C. Enders), 1:100 of mouse anti-OCT4 antibody OCT-3/4 (C-10; sc-5279, Santa Cruz Biotechnology Inc., Santa Cruz, California), 1:500 of mouse anti-PCNA antibody PCNA (PC10; sc-56, Santa Cruz Biotechnology Inc.); 1:50 of goat NGN3 (S-16; Sc-23832, Santa Cruz Biotechnology Inc.), and 1:50 of rat KI67 (M7249, DakoCytomation, Denmark). Incubation of the secondary antibody was performed at room temperature for 1 hour with a 1:400 dilution of fluorophor-conjugated secondary antibody (Alexa Fluor^R 488 goat anti-rabbit IgG, Molecular Probes Inc., Eugene, OR or Cy^™3-conjugated AffiniPure goat anti-rat IgG, Jackson ImmunoResearch Laboratories, West Grove, PA) or 1:100 dilution of HRP-conjugated goat anti-mouse secondary antibody (TSA™ Kit #3; T20913, Molecular Probes Inc.). For TSA signal amplification, tyramide labeling was performed following incubation with the HRP-conjugated secondary antibody according to the manufacturer’s recommendations. TOTO-3 iodide (T3604, Molecular Probes Inc.) counterstain was applied together with the secondary antibody and incubated for one hour. Samples were then washed 3 times in PBST (5 min. each) and mounted onto glass slides using Fluoromount-G (Southern Biotechnology Associates Inc., Birmingham, AL). Digital images were collected on either an Olympus (Olympus 1X71) inverted, a Nikon (Eclipse 80i) upright microscope, or a Leica inverted confocal (DMIRE2/TCS SP2 with argon/krypton and helium/neon lasers and excitation wavelengths 488, 543, and 594 nm) microscope and processed using Adobe Photoshop.

Western blot analysis

Western blot analysis was performed as described [28]. Briefly, tissue extracts from 17-days-old mice were resolved by SDS-polyacrylamide gel electrophoresis, transferred to PVDF, and probed overnight at 4°C with a 1:1000 dilution of DMRT1 antibody. Following incubation with secondary antibody (goat anti-rabbit antibody conjugated to horse radish peroxidase) protein complexes were visualized by chemiluminescence.

RESULTS

DMRT1 antibody generation and specificity

To evaluate cellular expression of DMRT1, a rabbit polyclonal antibody was developed against the carboxyl-terminal 15 amino acids of rat DMRT1 and used in immunohistochemical studies. DMRT1 antibody specificity was first evaluated by immunohistochemistry and western blot analysis. Analysis of testis sections from mice at 15.5 dpc revealed strong cross-reactivity with the DMRT1 antibody in the periphery of testis cords, which was absent in sections treated with antibody pre-absorbed with the peptide antigen or with preimmune serum (Figs. 1A & B and data not shown). Western blot analysis using the DMRT1 antibody showed specific recognition of a protein in lysates from MSC-1 cells (a mouse Sertoli cell line) transfected with a DMRT1 expression vector (Fig. 1C). Notably, the detected protein was absent from lysates derived from cells transfected with empty vector or when preimmune serum was used (Fig. 1C). Lysates from mouse tissues were also evaluated by western blot analysis, which revealed a similar-sized protein only in testis lysates (Fig. 1D). Importantly, in both transfected cells and tissue lysates, the detected protein resolved as a doublet and its position on the gel gave an estimated size of 60 kDa, which is significantly larger than that calculated from its amino acid sequence (~40 kDa). In summary, the analysis shows that the antibody is specific for DMRT1 and predicts, due to differences in the calculated and actual molecular mass, that DMRT1 is posttranslationally modified.

Figure 1. DMRT1 antibody specifically.

Immunohistochemical analysis of sequential sections from 15.5 dpc mouse testis using either A) DMRT1 immune serum or B) DMRT1 immune serum pre-adsorbed with its peptide antigen. C) Western blot analysis of lysates from MSC-1 cells transfected with a pcDNA3-based expression vector for rat DMRT1 (DMRT1) or empty pcDNA3 (vector), using the DMRT1 antibody (left) or pre-immune serum (right). D) Western blot analysis of tissue lysates from 17 day-old mice. Positions of the molecular weight markers (kDa) are indicated on the left. The membrane was stained with Ponceau S (left) prior to detection with the DMRT1 antibody (right). Both 10 and 20 micrograms of protein lysate (marked above the lane) were used for each indicated tissue. A and B; 20X magnification.

Sex-dependent differences in DMRT1 expression are not observed until after gonad differentiation

To evaluate DMRT1 expression during gonad development, immunohistochemistry was performed on embryos from 10.5 dpc through 15.5 dpc. At 10.5 dpc, before Sry (sex determining region Y) initiates testis formation, DMRT1 expression was restricted to the gonad in both sexes (Figs. 2A–C, data for XX embryo not shown). Through embryonic day 11.5, before any morphological difference was observed in the gonads, the pattern and intensity of DMRT1 staining was similar in XX and XY gonads (Figs. 2D & G). To help identify the DMRT1-expressing cells, gonads were stained for both DMRT1 and GCNA1, a germ cell-specific protein [29]. In XX and XY gonads at 11.5 dpc, the majority of immunoreactive cells were positive only for DMRT1 (green), indicating predominant somatic cell expression at this time. GCNA1 staining was also similar in both 11.5 dpc gonads, with each showing two populations of germ cells; one positive for both DMRT1 and GCNA1 (orange/yellow) and the other positive for only GCNA1 (red, Figs. 2F & I, insets; arrowheads mark DMRT1⁺/GCNA1⁺ cells, diamonds GCNA1⁺-only cells, arrows DMRT1⁺-only cells). Thus, no sexual differences were noted in the level or pattern of DMRT1, in either somatic or germ cell lineages at 11.5 dpc (Figs. 2D–I). However, by 12.5 dpc when the gonads have begun to differentiate, differences in DMRT1 expression were apparent between XX and XY gonads, both with respect to the location and nature of the immunoreactive cells. In the XY gonad, DMRT1 was localized to the testis cords, where it was observed in both Sertoli cells and germ cells (Figs. 2J–L). Notably, expression of DMRT1 was highest in Sertoli cells, as indicated by the intensity of green staining in cells negative for GCNA1 (Figs. 2J–L, insets; arrowheads mark germ cells and arrows Sertoli cells). In contrast, expression of DMRT1 in 12.5 dpc ovaries was highest in germ cells and only a few DMRT1⁺ somatic cells were evident, as indicated by the paucity of green stained cells that were negative for GCNA1 (Figs. 2M–O, inset; arrowheads mark germ cells and arrows Sertoli cells). Thus, sex-specific differences in DMRT1 expression were apparent by 12.5 dpc, with somatic cell levels of DMRT1 high in the testis and waning in the ovary.

Figure 2. DMRT1 and GCNA1 expression in mouse embryonic gonads prior to sexual differentiation.

Immunohistochemistry was used to examine DMRT1 and GCNA1 protein expression during gonad development. A) The DMRT1-specific antibody was used to reveal DMRT1 expression in 10.5 dpc XY embryo. Expression was detected by enzymatic staining for horse radish peroxidase. 4X magnification. B & C) Higher magnification of boxed region in A or B, respectively, showing expression in the genital ridge. Immunofluorescent detection of DMRT1 (green) and GCNA1 (red) in 11.5 dpc XY (D–F) and XX (G–I) gonads and 12.5 dpc XY (J–L) and XX (M–O) gonads. Insets show higher magnification with arrowheads indicating DMRT1⁺ germ cells, arrows DMRT1⁺ somatic cells, and diamonds DMRT1⁻ germ cells. Merged images (F, I, L, & O) show localization of DMRT1 in germ cells (orange/yellow). D–O were cropped from original 20X magnification.

DMRT1’s sexually-dimorphic profile in the embryo reflects differences in somatic cell expression

To further evaluate the sex-specific differences in DMRT1 expression, immunofluorescent staining for DMRT1 and GCNA1 was compared between the testis and ovary from 13.5 dpc through adult. Similar to the observed staining pattern at 12.5 dpc, DMRT1 expression at 13.5 dpc was dispersed throughout the ovary and predominantly localized with the GCNA1⁺ germ cells (Fig. 3 bottom, yellow/orange stained cells, arrowheads in insets). In the testis at 13.5 dpc, however, robust DMRT1 expression was detected in somatic cells, as indicated by the intense green staining in Sertoli cells located at the cord periphery (Fig. 3, top, arrows mark Sertoli cells and arrowheads germ cells). DMRT1 expression was also observed in germ cells, which now appeared more intensely stained than at 12.5 dpc. Thus by 13.5 dpc, increased germ cell expression was indicated in both sexes, while somatic cell expression was predominantly male-specific. By 15.5 dpc, DMRT1 in both XX and XY germ cells was nearly absent and the male-dominated expression in somatic cells even more evident (Fig. 4). In total, the data revealed that, during embryonic development, DMRT1 expression initiated similarly in somatic cells and germ cells of the XX and XY gonads, followed a similar course in the germ cells, and diverged significantly in somatic cells after 11.5 dpc. In both XX and XY germ cells, DMRT1 levels peaked around 13.5 dpc followed by its near absence by 15.5 dpc. In contrast, the pattern for DMRT1 was markedly different between XX and XY somatic cells, with its presence continuing throughout testis development and terminating in the ovary by 12.5 dpc.

Figure 3. DMRT1 expression in 13.5 dpc developing mouse gonads.

Enzymatic (left) and immunofluorescent detection of DMRT1 (green) and GCNA1 (red) in gonads from 13.5 dpc XY (top) and XX (bottom) embryos. Insets show higher magnification with arrowheads indicating DMRT1⁺ germ cells and arrows DMRT1⁺ somatic cells. Merged images (right) show localization of DMRT1 in germ cells (orange/yellow). 20X magnification.

Figure 4. DMRT1 expression in 15.5 dpc developing mouse gonads.

Enzymatic (left) and immunofluorescent detection of DMRT1 (green) and GCNA1 (red) in gonads from 15.5 dpc XY (top) and XX (bottom) embryos. Insets show higher magnification with arrowheads indicating DMRT1⁺ germ cells and arrows DMRT1⁺ somatic cells. Merged images (right) show localization of DMRT1 in germ cells (orange/yellow). Top, 20X magnification. Bottom, 32X magnification.

After birth, DMRT1 expression in the germ cells reappears in the testis but not the ovary

Examination of DMRT1 in postnatal testes and ovaries showed the sexually dimorphic expression was maintained after birth, with both immunohistochemical and western blot analysis demonstrating its continued expression in the testis and absence from the ovary (Figs. 1D, 5 & 6). DMRT1 expression in Sertoli cells was observed throughout postnatal testis development, while in germ cells, it reappeared in a limited number of gonocytes shortly after birth (0.5 dpp) and subsequently localized to germ cells located at the tubule periphery (Fig. 6). Following initial detection of DMRT1 in a limited number of gonocytes at 0.5 dpp, its expression extended to the remaining germ cells, until most or all were DMRT1⁺ by 7 dpp (Fig. 6). This time point also showed increased numbers of germ cells, predominantly located at the basal side of the tubule, and, with exception of one or two cells per tubule, significantly higher DMRT1 levels in germ cells than Sertoli cells. Coincident with the acquisition of more advanced spermatogenic cell types at 8 dpp was the appearance of DMRT1⁻ germ cells (red) and tubules with distinct cohorts of germ cells that could be distinguished by their DMRT1 expression pattern (Fig. 6, 8 dpp). Tubules at this stage were divided according to both germ cell DMRT1 levels and resident number of DMRT1⁺ germ cells. The noted patterns were designated H (abundant, robust-expressing germ cells & few or no DMRT1⁻ germ cells), M (abundant, moderately expressing germ cells) and L (no or few DMRT1+ germ cells and many DMRT1 negative germ cells; Fig. 6, 8 dpp). Distinct patterns of DMRT1-expressing germ cells were observed through 10 and 20 dpp, although the abundance of a specified germ cell pattern appeared to vary (Fig. 6, 10 and 20 dpp). Notably, by 20 dpp, the number of tubules containing few or no DMRT1⁺ germ cells (L) increased significantly, a feature that remained as the testes continued their maturation and the DMRT1⁺ germ cells further restricted to the basement membrane (Fig. 6 and data not shown). In addition to the changes observed in germ cells, the expression profile revealed varied DMRT1 levels among the Sertoli cell population just after birth (0.5 and 2 dpp), which largely normalized by 7 dpp (Fig. 6).

Figure 5. DMRT1 expression during postnatal development of the mouse ovary.

Enzymatic (left) and immunofluorescent detection of DMRT1 (green, 2^nd column) and GCNA1 (red, 3^rd column) individually, or as merged images (far right column) in ovaries isolated from newborn (P 0.5), 17 days postpartum (17dpp), and adult mice. No specific staining for Dmrt1 was detected in images for Dmrt1, either alone (2^nd column from left) or merged with GCNA (right). 0.5 dpp, 32X magnification. 17 dpp, 20X magnification. Adult enzymatic 10X magnification. Adult immunofluorescence, 6.4X magnification.

Figure 6. DMRT1 expression during postnatal development of the mouse testis.

Dual immunofluorescent staining for DMRT1 (green) and GCNA1 (red) in testes isolated from 0.5 days postpartum (dpp), 2 dpp, 7 dpp, 8 dpp, 10 dpp, 20 dpp, 35 dpp, and adult mice. Merged images show DMRT1 expressed in Sertoli cells (green) and co-expressed with the germ cell marker GCNA (orange/yellow cells) at all stages of postnatal testis development. At pre-pubertal stages from 8 to 20 dpp, individual tubules were designated according to the number of positive germ cells and their DMRT1 levels. The noted patterns were designated H (abundant, high-expressing and few negative cells), M (abundant, moderately-expressing cells) and L (predominantly DMRT1⁻germ cells). 0.5 dpp, 40X magnification. All others 20X magnification.

Figure 8. DMRT1 and OCT4 expression during postnatal development of the mouse testis.

DMRT1 (green) and OCT4 (blue) were visualized by dual immunofluorescence and DNA by TOTO-3 counter-staining (white/grey) in testes isolated from 0.5 dpp, 1 dpp, 2 dpp, 7 dpp, 10 dpp, and adult mice. Merged images show DMRT1⁺/OCT4⁺ germ cells (aqua) at all ages examined. DMRT1 (red) and NGN3 (green) were also visualized by dual immunofluorescence in testes from 7 dpp mice, revealing DMRT1⁺/NGN3⁺ germ cells. DMRT1⁻/OCT4⁺ (dark blue) or DMRT1⁻/NGN3⁺ (green) cells were infrequent or not observed at the time points tested. 0.5 dpp, 2dpp, & 7dpp are 40X; 10 dpp, & adult 60X; Ngn3/7 dpp 20X.

DMRT1 is expressed in mitotically active gonocytes and spermatogonia

As revealed by its expression pattern, DMRT1’s presence in germ cells closely associated with their mitotic activity. That is, DMRT1 was lost from embryonic germ cells at a time when PGCs exit the cell cycle, either to arrest in mitosis (male) or enter meiosis (female), and reemerged in testicular germ cells at a time when gonocytes begin their exit from cell cycle arrest [27;30–42]. To more accurately assess its association with mitotically active germ cells, DMRT1 immunohistochemistry was performed together with KI67 or PCNA, which detects all phases of the cell cycle except G0 and, in the testis, types A, I, and B spermatogonia [21–24]. At 0.5 dpp, mitotically active cells, as revealed by KI67 (red), were located within the tubule periphery and interstitium (Fig. 7). The morphology and location of DMRT1⁺/KI67⁺ cells indicated most or all were Sertoli cells, as the gonocytes at 0.5 dpp were still centralized within the tubules and negative for KI67 (Fig. 7). Immunofluorescent staining of PCNA (blue), DMRT1 (green), and GCNA (red) showed mitosis initiated between 1 and 2 dpp in germ cells and closely correlated with their acquisition of DMRT1, as nearly all PCNA⁺/GCNA⁺ cells (proliferating gonocytes) were also DMRT1⁺ (Fig. 7 and data not shown). By 7 dpp, all observed germ cells (GCNA⁺) were DMRT1⁺, PCNA⁺, and localized to the tubule periphery (Fig. 7). PCNA staining at 8 dpp continued to mark all or most GCNA⁺ germ cells, which had segregated into two populations; DMRT1⁺ cells located at the periphery and DMRT1⁻ cells located more centrally (Fig. 7). DMRT1-expressing germ cells retained their basal localization through adulthood and many, but not all, DMRT1⁺ germ cells were also PCNA⁺ (Fig. 7 and data not shown). In the adult testis, KI67 staining (red) localized predominantly to germ cells within the tubules, where it overlapped extensively with cells expressing DMRT1 (green). Thus, most KI67⁺ adult germ cells were also DMRT1⁺ but there were notable differences in DMRT1 levels, reflected in the shade of the merged image (Fig. 7). Adult tubules also exhibited a large variation in the pattern of proliferating, DMRT1⁺ germ cells, with some tubules containing numerous KI67⁺/DMRT1⁺ germ cells, while others had few or no detectable KI67⁺ cells.

Figure 7. DMRT1 expression in mitotically active germ cells of the postnatal developing mouse testis.

Dual immunofluorescent staining was performed for DMRT1 (green) and KI67 (red) in testes isolated from 0.5 dpp and adult mice. Likewise, triple immunofluorescent staining was performed for DMRT1 (green), GCNA (red), and PCNA (blue) in testes isolated from 2 dpp, 7 dpp, 8 dpp, and 10 dpp. Shown are merged images, which indicate DMRT1 in proliferating Sertoli cells at 0.5 dpp (yellow/orange) through 8 dpp (light blue), and in proliferating germ cells from 2 dpp through adult. In merged images of 2 dpp, 7 dpp, 8 dpp, and 10 dpp testes, DMRT1⁺ and DMRT1⁻ proliferating germ cells project as white/light pink and dark pink cells, respectively. In the adult, DMRT1⁺ proliferating germ cells are depicted as yellow/orange cells in the merged images. 0.5 dpp, 40X magnification. Others, 20X magnification.

DMRT1 is expressed in OCT4 and NGN3 expressing spermatogonia

The distribution pattern of DMRT1 indicated it resides within the initial spermatogonia population that arises from gonocytes during the first week after birth (Figs. 6 and 7). To further evaluate the nature of DMRT1-expressing spermatogonia, OCT4 and NGN3, markers of undifferentiated spermatogonia, including the stem cells, were examined together with DMRT1 [25–27]. Consistent with previously published studies, OCT4 expression persisted in all gonocytes after birth and then, with further development, became restricted to a limited number of germ cells on the basement membrane (Fig. 8). Co-expression of DMRT1 and OCT4 was first observed with the initiation of DMRT1 shortly after birth (0.5 dpp) and continued through adulthood (Fig. 8). By 2 dpp and at each developmental stage thereafter, DMRT1 was observed in all OCT4⁺ germ cells, in addition to germs cells negative for OCT4 (Fig. 8). Similarly, DMRT1 (red) was noted in all NGN3⁺ (green) cells as well ones negative for NGN3 (Fig. 8, bottom right; [27]). Together the studies suggest DMRT1 is present within both stem and proliferating spermatogonia.

DISCUSSION

Mammalian sex determination is often described as the sequential progression of three interdependent phases, characterized by the acquisition of genetic, gonadal, and phenotypic sex [43;44]. The first phase occurs upon fertilization, when genetic sex is specified by donation of either an X or Y chromosome from sperm. The genetic sex dictates the developmental fate of the gonad (gonadal sex), with the determining factor being the presence or absence of a gene on the Y chromosome, Sry, which acts in a dominant manner to induce testis formation [44–51]. In the third phase, the developing testis or ovary determines the accompanying endocrine environment, which, in turn, drives development of the secondary sex characteristics and thus phenotypic sex [52]. While each of these stages is critical to the final sexual outcome, the pinnacle feature of vertebrate sex determination is gonad development [43].

Early gonad development is characterized by the formation of a bipotential structure, comprised of primordial germ cells and somatic precursor cells, and is identical in males and females and independent of genetic sex [23;32–35]. The developing gonad remains indifferent until it commits to either a testicular or ovarian pathway and sexual identity is determined. Commitment to the testicular pathway occurs with the transient induction of Sry in Sertoli cell precursors, which induces their differentiation and subsequent ability to orchestrate testis organogenesis. In mice, approximately 36 hours after induction of Sry, or 12.0 to 12.5 dpc, testis differentiation is clearly visible and characterized by the appearance of cord-like structures formed by the assembly of Sertoli cells around the primordial germ cells. In the absence of Sry, the gonad commits to the ovarian pathway and no obvious signs of differentiation are detected until birth, when follicular structures become visible. In additional to these distinct morphological differences, development of the testis and ovary diverge in terms of germ cell participation, as testis development occurs normally in the absence of germ cells, while in the ovary, germ cells are needed to organize and maintain its structure [38;53–60]. The earliest recognizable germ cell precursors, the primordial germ cells (PGCs), originate from cells in the proximal epiblast, under the inductive influence of bone morphogenetic proteins, and migrate along the hind gut to populate the gonads around 10.0 dpc [41]. Once in the indifferent gonad, PGCs proliferate until 13.5 dpc, at which time they commit to become either oocytes or prospermatogonia, a decision that depends on the sex of gonadal somatic cells [41;59;61;62]. In the ovary, PGCs progress through the first meiotic prophase and arrest in diplotene just after birth, while in the testis, meiotic entry is blocked and male germ cells arrest in mitosis at G0/G1 until just after birth when they resume proliferation [38;41;63;64].

Expression profiles and mutant animal models have unveiled numerous genes important to gonad development and sex determination and, collectively, show that many participants in sex determination have a strong preference towards testis development and cellular expression profiles that stress the importance of Sertoli cell differentiation and expansion [43]. In developing germ cells, studies have emphasized the relevance of gene expression profiles that diverge between migrating germ cells and those residing in the gonads, with the latter characterized by enhanced expression of several germ cell-specific proteins, including mouse vasa homolog, GCNA1, DAZL, OCT4, NANOG, and STELLA [41;65]. In the current study, immunohistochemical analysis was used to track the expression profile of DMRT1, a protein implicated in both Sertoli cell and germ cell function, but whose cellular program had not been firmly established. Hence, DMRT1 was traced throughout embryonic and postnatal development to help predict its activity and regulation and provide additional insights on its potential role in sex determination and germ cell development.

Comparison of DMRT1 and GCNA1, which is specific to germ cells from 11.5 dpc through the dictyate/late diplotene stage of meiotic prophase I, showed that gender differences in DMRT1 resulted from distinct cellular patterns that were both sex-specific and developmentally dynamic. Prior to any distinction in gonad morphology (i.e. 10.5 & 11.5 dpc), DMRT1 expression was similar in males and females and observed in both somatic cells and germ cells. Coincident with the first signs of testis differentiation at 12.5 dpc, DMRT1 localized to the testis cords, where it predominated in Sertoli cells, relative to the low-expressing germ cells. In contrast, DMRT1 expression in 12.5 dpc ovaries was markedly lower in somatic cells than germ cells (Figs. 2J & M insets, compare somatic and germ cell signal intensities for DMRT1). Thus, the comparison revealed a sexual preference for DMRT1 in male somatic cells, which coincided with the morphological changes that mark sexual differentiation. This preference continued throughout embryonic and postnatal development, with DMRT1 evident in Sertoli cells at all ages examined and lacking in ovarian somatic cells from 13.5 dpc onward. Notably, DMRT1 was not detected in any ovarian cell type after birth, either by immunohistochemistry or western blot analysis. While these findings are consistent with DMRT1’s male-specific role, they are in conflict with a previous report showing DMRT1 expression in ovarian granulosa, theca, and stromal cells of adult mice [66]. The reason for this discrepancy is unclear but may reflect differences in the sensitivity or specificity of the employed antibodies.

DMRT1’s preferred expression in Sertoli cells at 12.5 dpc is consistent with a role in sex determination and testis differentiation that occurs subsequent to the dominant actions of sex determining factors, Sry and Sox9 (Sry-related HMG box containing gene 9). In mammals, however, DMRT1’s activity during sex determination and early testis development is equivocal, due to conflicting evidence from humans and mice. In support of its role in sex determination, deletions of distal portions of human chromosome 9 (e.g., 9p24.3), which contains DMRT1 and/or its upstream region, were observed in patients with 46, XY sex reversal [67–69]. In mice, however, ablation of Dmrt1 did not reveal a role in sex determination, as embryonic gonadogenesis and thus sex determination was normal [18]. Instead, severe defects were observed in postnatal testes and disclosed Dmrt1’s importance in Sertoli differentiation and male germ cell survival. The delayed phenotype in mice suggested a potential genetic redundancy between Dmrt1 and other gene(s), such as Dmrt3, which is located immediately 3′ to Dmrt1 in mouse and human [68]. Indeed, based on its chromosomal position, DMRT3 was implicated, along with DMRT1, in XY sex reversal caused by deletions in human chromosome 9p24.3 [70]. Furthermore, analysis of other DM domain genes during mouse development showed some expressed in the embryonic gonads [19]. Notably, Dmrt3 was present at similar levels in indifferent XX and XY gonads and decreased in the ovary by 13.5 dpc, resulting in proportionally higher levels in the testis [19]. While Dmrt3 was found in a wide range of tissues, its Dmrt1-like expression in the gonad suggests it functions in support of Dmrt1 during gonadogenesis and offers one potential explanation for the discrepancy between DMRT1’s ascribed functions in humans and mice.

DMRT1 expression in germ cells was remarkably similar in embryonic testes and ovaries but differed after birth. Both XX and XY PGCs had moderate levels of DMRT1 at 11.5 and 12.5 dpc. Levels then rose slightly by 13.5 dpc, followed by near extinction at 15.5 dpc. After birth, DMRT1 was again observed in the germ cells, but only in the testis and only in premeiotic cells. What is most noteworthy about the dynamics of DMRT1 expression in these germ cells is its correlation to their mitotic activity. Thus, loss of DMRT1 from embryonic germ cells corresponded to times when PGCs exit the cell cycle, either to arrest in mitosis (male) or to enter meiosis (female) [37–42]. The reemergence of DMRT1 in postnatal testicular germ cells occurred at a time when gonocytes begin their exit from cell cycle arrest and transition into one of two distinct spermatogonia populations; C-KIT⁺ differentiating spermatogonia and NGN3⁺ undifferentiated spermatogonia [27;30–36]. Association of DMRT1 with germ cell mitotic activity was more firmly established with the demonstration that DMRT1 resided in most or all proliferating (either PCNA or KI67 positive) spermatogonia of developing and adult testes. The presence of DMRT1 in mitotically active germ cells and its segregation from the more advanced spermatogenic cells that emerge at 8 dpp suggest DMRT1 is involved in the initial phase of germ cell proliferation and expansion of undifferentiated spermatogonia [35;71;72]. Co-localization studies with OCT4 and NGN3 indicated DMRT1 functions within spermatogonial stem cells, in addition to germ cells that diverge from these populations, presumably undifferentiated-proliferating as well as differentiating type A spermatogonia. Thus, shortly after birth DMRT1 co-localized with OCT4 and NGN3, some of which (NGN3⁺/OCT4⁺) become the stem cell population, while others the differentiating spermatogonia, and DMRT1 is highly expressed within these cells as they proliferate to expand the initial spermatogonia cell populations.

In summary, DMRT1’s expression profile and evidence from the literature indicate distinct roles in somatic and germ cell lineages. In somatic cells, DMRT1 is implicated in differentiation of Sertoli cells. Thus, soon after Sry and Sox9 initiate Sertoli cell differentiation and testis morphogenesis is observed, there is a notable, male-specific retention of DMRT1 in somatic cells, suggesting its continued expression is required to complete or maintain Sertoli cell differentiation. In germ cells, DMRT1 levels correlated with mitotic activity, implicating it in cell cycle control and proliferation. This emerging image of DMRT1 is supported by its deletion from the mouse, which revealed an early block in spermatogenesis and poorly differentiated Sertoli cells. [18]. In these animals, testis structure was normal through 7 dpp but, by 10 dpp, germ cell numbers dropped significantly, a result attributed to decreased cell proliferation and supported by recent studies showing defective germ cell mitosis at 7 days postpartum [18;73]. However, because DMRT1 is expressed in both Sertoli cells and germ cells, which are functionally dependent on each other, it has not been possible to ascribe any cell-specific functions to DMRT1. Thus, it is hoped that conditional deletions of DMRT1 will soon resolve its distinct cellular roles.