The DEAD-box helicase DP103 (Ddx20, Gemin3) is a multifunctional protein that interacts with Epstein-Barr virus nuclear proteins (EBNA2/EBNA3) and is a part of the spliceosomal small nuclear ribonucleoproteins complex. DP103 also aggregates with the micro-RNA machinery complex. We have previously shown that DP103 interacts with the nuclear receptor steroidogenic factor-1 (SF-1, NR5A1), a key regulator of reproductive development, and represses its transcriptional activity. To further explore the physiological function of DP103, we disrupted the corresponding gene in mice. Homozygous Dp103-null mice die early in embryonic development before a four-cell stage. Although heterozygous mice are healthy and fertile, analysis of steroidogenic tissues revealed minor abnormalities in mutant females, including larger ovaries, altered estrous cycle, and reduced basal secretion of ACTH. Our data point to diverse functions of murine DP103, with an obligatory role during early embryonic development and also in modulation of steroidogenesis.
THE NUCLEAR RECEPTOR steroidogenic factor-1 (SF-1, NR5A1) is essential for development of steroid-producing glands and the ventromedial hypothalamus. SF-1 also controls the expression of diverse enzymes involved in the biosynthesis of steroid hormones and coregulates the function of pituitary gonadotropes (reviewed in Refs. 1–3). Inactivating mutations in the Sf-1 gene cause sex reversal and severe endocrine dysfunction in rodents and humans. Akin to other members of the nuclear receptor family, SF-1 harbors a conserved DNA-binding domain and a ligand-binding domain. Although considered an orphan receptor, discrete phospholipids were recently found to bind SF-1 and modulate its transcriptional activity (4, 5). The activity of SF-1 is also modulated by posttranslational phosphorylation (6) as well as by direct protein-protein interactions with diverse proteins, including DAX-1, SOX9, WT1, GATA-4, Ptx1, ZIP67, GIOT-1, c-Jun, SNURF, and RIP140 (7–17).
We previously identified a discrete domain within SF-1 that imparts transcriptional repression through the recruitment of DEAD-box protein-103 (DP103)/Gemin3/Ddx20, a member of the DEAD-box family of putative RNA helicases (18, 19). We confirmed that DP103 acts as transcriptional repressor of SF-1, and this repression requires the C-terminal region that is also germane for DP103’s helicase activity (19). Additionally, DP103 causes sumoylation of SF-1 and relocalization to discrete nuclear foci (20). DP103 also represses the transcriptional activity of Egr2 (21) and FOXL2 (22) and mediates the transcriptional repression of the N-terminal repression domain of METS and both N- and C-terminal repression domains of ERF via recruitment of histone deacetylase proteins (23, 24). Finally, consistent with the notion that RNA helicases are involved in multiple aspects of RNA metabolism, DP103 associates with the survival motor neuron (SMN) protein, which is required for the assembly of spliceosomal small nuclear ribonucleoproteins (25, 26) (reviewed in Ref. 27).
Although some of the molecular pathways of DP103 are known, little is known about the role of DP103 in development and physiology. To gain insight into the function of DP103 in vivo, we used homologous recombination to inactivate the Dp103 locus in mice. We found that homozygous Dp103-null mutant embryos die before implantation. Dp103 heterozygous mutant mice are viable and fertile. Further characterization of Dp103 heterozygotes with a primary focus on reproduction revealed several alterations in the endocrine system, consistent with the influence of DP103 on SF-1.
Materials and Methods
Mice and genotyping
All experiments were approved by the Washington University Animal Studies Committee. Mice were maintained in the Washington University barrier animal facility. To create our transgene, a 7.5-kb fragment of the mouse Dp103 gene that contains exon 1 was excised from a 129/Sv BAC plasmid and subcloned into pSP72 (Promega, Madison, WI). The resultant plasmid was used as a recipient for recE/T-mediated recombination in Escherichia coli as described by Stewart and colleagues (28). The donor-targeting fragment for bacterial recombination included a disruptive knock-in cassette made of a 2.8-kb PCR product, generated from a plasmid harboring myc-tagged enhanced green fluorescent protein (EGFP) upstream of a loxP-flanked neomycin/kanamycin resistance cassette with a dual eukaryotic/prokaryotic (PGK/Tn5) promoter. This cassette was flanked by 66-bp arms that were homologous to Dp103. Recombinants were selected using appropriate antibiotics and confirmed by PCR and Southern blotting. The linearized vector was electroporated into 129/Sv embryonic stem (ES) cells, and two independent positive clones were obtained after selection with G418 and confirmed by Southern blotting with a probe outside the homology arms. The two ES clones were microinjected into C57BL/6 blastocysts to obtain chimeric mice. Resulting chimeras were bred with C57BL/6 mice, and offspring derived from ES cells were identified by their coat color, Southern blotting, and PCR of tail DNA. Genotyping was carried out by PCR using the following three primers: 5′-TTGACAACCTCCGAGTCCACTATGGCT, 5′-TTCCTCGTGCTTTACGGTATCGCCGCT, and 5′-ACTGATGCCACCAGATACCAAATA. For PCR screening of embryonic genotype, we used a DNA extraction kit (Epicenter Biotechnologies, Madison, WI). DNA was extracted from each embryo in 12 μl DNA extraction solution and processed following the manufacturer’s instructions.
Embryo harvest
Female mice heterozygous for Dp103 were generated after back-crossing into C57BL/6 background mice and were superovulated using ip injections of 5 IU equine gonadotropin, followed 48 h later by 5 IU human chorionic gonadotropin (both from Sigma Chemical Co., St. Louis, MO). The injected female mice were then placed with Dp103+/− males of proven fertility and checked the next morning for a copulation plug. Timing after human chorionic gonadotropin injection and mating was used to recover one-cell embryo (24 h), two-cell embryo (48 h), or morula (72 h). Embryos were flushed from dissected oviducts that were attached to the proximal end of the uterine horn under stereomicroscopy (Nikon SMZ-10A) using a blunted 30-gauge needle. The embryos were placed in warmed human tubal fluid medium (Irvine Scientific, Irvine, CA) supplemented with 0.25% BSA and 0.3 mg/ml hyaluronidase (Sigma) to remove the cumulus cells in one-cell embryos. Flushed embryos were placed on Superfrost Slides (Fisher Scientific, Waltham, MA), fixed in 3% paraformaldehyde for 20 min at room temperature, and processed as described. At least 10 pregnancies and more than 80 embryos were examined for Dp103−/− genotype at the different gestational ages.
DP103 antibodies, immunofluorescence, and immunohistochemistry
The production of anti-DP103 polyclonal antibodies was previously described (29). The antibody was affinity purified using an AminoLinkPlus kit (Pierce, Rockford, IL) and confirmed by electrophoresis. For detection of embryonic DP103 protein, 1- to 3-d embryos were permeabilized for 10 min in 0.1% Tween 20 in PBS. Fixed embryos were blocked for 1 h in 20% normal donkey serum, 0.5 mg/ml purified antimouse CD16/CD32 (BD Biosciences, San Jose, CA), and 2% BSA. The sections were incubated with anti-DP103 (stock concentration is 2.3 μg/μl, diluted to a working concentration of 1:100) or goat anti-GFP antibody (1:50; Santa Cruz Biotechnology, Santa Cruz, CA). After washing three times in PBS with 2% BSA, the specimens were incubated with antirabbit Alexa 594 or antigoat Alexa 488 (Invitrogen, Carlsbad, CA). Sections were counterstained with TOPRO-3 iodide (Invitrogen) at a concentration of 4 μm, washed three times in PBS, and mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA).
For histological assessment of DP103 expression, ovaries, testes, and adrenal glands were dissected from wild-type or Dp103+/− mice (6–12 wk old, virgins), weighed, and fixed in 4% paraformaldehyde in PBS overnight at 4 C. Sections were cut from paraffin-embedded tissues, mounted on Superfrost-plus slides, and after deparaffinization were subjected to antigen retrieval using 0.01 m sodium citrate (pH 6.0) in a microwave. After washes, the sections were blocked using 10% goat serum, 0.3% Triton X-100, and 1% BSA in Tris-buffered saline for 1 h. Rabbit anti-DP103 was diluted 1:100 in blocking buffer and incubated on sections overnight at 4 C. Fluorescent sections were washed and incubated with Alexa 488 antirabbit antibody (1:400; Invitrogen) for 1 h. Nuclei were stained using TOPRO-3 iodide (Invitrogen). Preimmune serum or absent primary antibody was used to confirm the specificity of the DP103 antibody. Ovarian follicles were analyzed using a single slide in a longitudinal midsection of each ovary by a researcher who was blinded to the genotype. After application of the secondary antibody, some ovarian sections were incubated with the ABC reagent (Vector) for 30 min and immunoreactivity visualized using AEC chromogen (Vector).
RT-PCR
Total RNA was isolated using Tri reagent (Sigma) following the manufacturer’s instructions, then incubated with DNase I (1 U/10 μg RNA; Ambion, Austin, TX) at 37 C for 20 min. For quantitative PCR, we initially performed RT with 0.5 μg total RNA in a 50-μl RT reaction that included 5 μl TaqMan RT buffer, 11 μl 25 mm MgCl2, 10 μl 10 mm dNTP mixture, 2.5 μl 50 μm random primer, 20 U RNase inhibitor, and 1.25 μl murine leukemia virus reverse transcriptase (Applied Biosystems, Foster City, CA). Transcript levels were determined by quantitative PCR using Geneamp 5700 or 7300 Sequence Detection System (Applied Biosystems). Each 50 μl of PCR included 3 μl RT products, 300 nm each of forward and reverse primers, and 25 μl SYBR-Green master mix (Applied Biosystems). Samples were normalized to parallel reactions using primers specific for 18S RNA as we previously described (30). The fold increase relative to control cultures was determined by the 2−ΔΔCT method (31). PCR amplification was performed with the following primers pairs: Cyp11a1, 5′-CCAGTGTCCCCATGCTCAA-3′ and 5′-CAGCTGCATGGTCCTTCCA-3′; StAR, 5′-GGAGATGCCGGAGCAGAGT-3′ and 5′-GCCAGTGGATGAAGCACCAT-3′; Cyp17a1, 5′-CCTGCACTTGGCAATACATCTG-3′ and 5′-AGGTCCCTCAACGTGGCATA-3′; Cyp19a1, 5′-CCACAGCTGAGAAACTGGAAGA-3′ and 5′-TCGTCAGGTCTCCACGTCTCT-3′; Cyp21a1, 5′-CCCGGGTTTTCTGCACTTC-3′ and 5′-TGTAGATGGGCCCGAGTTTC-3′; and 18S, 5′-CACGGCCGGTACAGTGAAA-3′ and 5′-AGAGGAGCGAGCGACCAA-3′.
Examination of the estrous cycle
The estrous cycle stage was determined by daily assessment of vaginal cytology over a period of 21 d. Animals were individually housed and given food and water ad libitum. Vaginal smears from four wild-type and four Dp103 heterozygous female mice (age 9–12 wk) were collected at 1000 h by investigators blinded to the mouse genotype. Scrapings were obtained using small cotton swabs prewetted with PBS, smeared onto a glass slide, fixed in absolute methanol, and stained with Giemsa (Sigma). The estrous cycle stage was determined by the presence of nucleated epithelial cells (NEC), cornified epithelial cells (CEC), and polymorphonuclear cells (PMN), with estrus enriched for CEC and some NEC without PMN, metestrus characterized by CEC and PMN, diestrus enriched for PMN with few NEC, and proestrus enriched mainly by NEC and few CEC and PMN. Ambiguous cycle status was ascertained based on determinations on the preceding and subsequent day and labeled as an intermediate value (32).
Restraint test and hormone measurement
Hormone levels were measured in the morning (2 h after lights on) and evening (1 h before lights off) serum samples as well as after morning restraint stress, performed by placing the mice in a 50-ml Falcon tube for 30 min as described by (33). Blood was collected from stressed and control mice, and plasma was separated by centrifugation and stored at −80 C until assay. Plasma concentrations of corticosterone (ICN, Costa Mesa, CA) and ACTH (Diasorin, Stillwater, MN) levels were measured as previously described (33).
Statistics
Each experiment was repeated at least three times. Significance was analyzed by Student’s t test, ANOVA with post hoc Bonferroni test, or χ2 test where appropriate and detailed in each figure legend (Primer of Biostatistics software package; McGraw-Hill, Inc., New York, NY). A value of P < 0.05 was determined significant.
Results
Expression of DP103 in steroid-producing tissues
Because of the prominent expression of DP103 in steroid-producing organs and the interaction of DP103 with SF-1 (18–20), we focused our attention on DP103 expression in these tissues (Fig. 1). Using immunofluorescent staining of the ovary, we detected strong expression of DP103 in follicular oocytes. There was no DP103 signal when preimmune serum was used instead of the primary antibody (Fig. 1, A and B). Within the adult testis, we detected DP103 protein in seminiferous tubules mainly in spermatocytes and spermatids (Fig. 1C) with a lower expression level in the interstitium. In the adrenal gland, we detected DP103 in the medulla and cortex, with a high level of expression in the zona fasciculata (Fig. 1D). We noted that the pattern of expression of DP103 only partly overlapped with that of SF-1, as previously described (34–38). We also confirmed DP103 expression at multiple stages of follicular development, with expression in the surrounding granulosa cells and in the corpora lutea, albeit at a slightly lower level compared with the oocyte (Fig. 1E).
Expression and localization of DP103 in steroid-producing organs. Immunofluorescence was performed for DP103 (green) and TOPRO-3 iodide for nuclear counterstaining (blue) as described in Materials and Methods. A, Mouse ovarian sections stained with preimmune rabbit serum, serving as a negative control. B, Mouse ovarian sections. Arrow indicates primary follicle; arrowhead indicates antral follicle. CL, Corpus luteum. C, Mouse testicular sections. D, Female adult mouse adrenal sections showing adrenal medulla (M) cortex (C), and x-zone (X). E, Expression of DP103 in follicular stages, including (from left) negative control, secondary follicle, antral follicle, and postrupture follicle, detected using immunohistochemistry as described in Materials and Methods. Data shown are from one experiment, which represents at least three repeats, each with qualitatively similar results. Scale bar, 50 μm.
Generation of Dp103 mutant mice
To elucidate the function of DP103, we inactivated Dp103 gene by targeted gene disruption. Our construct was designed to replace the genomic fragment containing the first exon of the Dp103 gene with a disruptive myc-EGFP-neo cassette using homologous recombination (Fig. 2, A and B). The myc-tagged EGFP sequences were inserted in frame with the authentic ATG start codon of Dp103 to mimic its expression in the deleted allele. We generated and examined two independent lines of Dp103 mutant mice, which yielded a similar phenotype. The male and female heterozygous offspring appeared phenotypically indistinguishable from wild-type littermates and exhibited similar fertility when observed for approximately 12 months. Whereas litter size was not significantly affected by the parental phenotype (Table 1), no homozygous mutants were found in the progeny of heterozygous intercrosses, suggesting that the complete loss of DP103 results in embryonic lethality. To determine the time of lethality of Dp103-null embryos, we determined the genotype of postimplantation embryos at different stages of development, collected from timed matings, but found no Dp103−/− mutants. To identify Dp103-null embryos in the preimplantation period, we took advantage of the fact that the disruptive Dp103 cassette harbored EGFP coding sequences. By monitoring the expression of DP103 and EGFP proteins, we identified EGFP-positive/DP103-negative embryos, indicative of Dp103−/− genotype, up to the two-cell stage (Fig. 2C). These data indicated that Dp103−/− mutant embryos could not survive beyond the early postzygotic period.
Targeted disruption of mouse Dp103 (Ddx20/Gemin3) gene. A, Schematic maps of the targeting vector, wild-type allele, and targeted allele. The filled boxes and solid lines between them represent exons and introns, respectively. The middle drawing represents the 5′ end of mouse Dp103 gene with the boxes labeled E1 and E2 representing exons 1 and 2, respectively. The targeting vector was linearized by PmeI. Restriction sites were ClaI (C), PstI (P), and SpeI (S). Neo, Neomycin resistance cassette. B, A representative Southern blot analysis (reproduced more than 10 times) of DNA extracted from wild-type and heterozygote mice for Dp103. Genomic DNA was digested with SpeI. The wild-type and targeted alleles yield 6.8- and 9.2-kb fragments, respectively (internal probe indicated; an external probe was used for ES cell screening). C, Detection of DP103 (red) and GFP (green) proteins in two-cell mouse embryos. The zygotes in the three panels were stained for both DP103 and GFP, demonstrating (from left) Dp103 wild-type, null, and heterozygous zygotes. Images shown are from one experiment, which represents four repeats, each with similar results.
Genotype analysis of matings from wild-type and Dp103 heterozygous mice
| Parental genotype | Pups/litter (mean ± sd) | Pup gender | Total | +/+ n (%) | +/− n (%) |
|---|---|---|---|---|---|
| F +/− M +/− (n = 73) | 6.8 ± 2.9 | All | 485 | 169 (34.8) | 316 (65.2) |
| M | 241 | 74 (30.7) | 167 (69.3) | ||
| F | 244 | 95 (38.9) | 149 (61.1) | ||
| F +/+ M +/− (n = 18) | 6.9 ± 2.6 | All | 79 | 42 (53.2) | 37 (46.8) |
| M | 40 | 22 (55.0) | 18 (45.0) | ||
| F | 39 | 20 (51.3) | 19 (48.7) | ||
| F +/− M +/+ (n = 22) | 7.9 ± 4.7 | All | 177 | 92 (52.0) | 85 (48.0) |
| M | 86 | 42 (48.8) | 44 (51.2) | ||
| F | 91 | 50 (54.9) | 41 (45.1) |
| Parental genotype | Pups/litter (mean ± sd) | Pup gender | Total | +/+ n (%) | +/− n (%) |
|---|---|---|---|---|---|
| F +/− M +/− (n = 73) | 6.8 ± 2.9 | All | 485 | 169 (34.8) | 316 (65.2) |
| M | 241 | 74 (30.7) | 167 (69.3) | ||
| F | 244 | 95 (38.9) | 149 (61.1) | ||
| F +/+ M +/− (n = 18) | 6.9 ± 2.6 | All | 79 | 42 (53.2) | 37 (46.8) |
| M | 40 | 22 (55.0) | 18 (45.0) | ||
| F | 39 | 20 (51.3) | 19 (48.7) | ||
| F +/− M +/+ (n = 22) | 7.9 ± 4.7 | All | 177 | 92 (52.0) | 85 (48.0) |
| M | 86 | 42 (48.8) | 44 (51.2) | ||
| F | 91 | 50 (54.9) | 41 (45.1) |
F, Female; M, male.
Genotype analysis of matings from wild-type and Dp103 heterozygous mice
| Parental genotype | Pups/litter (mean ± sd) | Pup gender | Total | +/+ n (%) | +/− n (%) |
|---|---|---|---|---|---|
| F +/− M +/− (n = 73) | 6.8 ± 2.9 | All | 485 | 169 (34.8) | 316 (65.2) |
| M | 241 | 74 (30.7) | 167 (69.3) | ||
| F | 244 | 95 (38.9) | 149 (61.1) | ||
| F +/+ M +/− (n = 18) | 6.9 ± 2.6 | All | 79 | 42 (53.2) | 37 (46.8) |
| M | 40 | 22 (55.0) | 18 (45.0) | ||
| F | 39 | 20 (51.3) | 19 (48.7) | ||
| F +/− M +/+ (n = 22) | 7.9 ± 4.7 | All | 177 | 92 (52.0) | 85 (48.0) |
| M | 86 | 42 (48.8) | 44 (51.2) | ||
| F | 91 | 50 (54.9) | 41 (45.1) |
| Parental genotype | Pups/litter (mean ± sd) | Pup gender | Total | +/+ n (%) | +/− n (%) |
|---|---|---|---|---|---|
| F +/− M +/− (n = 73) | 6.8 ± 2.9 | All | 485 | 169 (34.8) | 316 (65.2) |
| M | 241 | 74 (30.7) | 167 (69.3) | ||
| F | 244 | 95 (38.9) | 149 (61.1) | ||
| F +/+ M +/− (n = 18) | 6.9 ± 2.6 | All | 79 | 42 (53.2) | 37 (46.8) |
| M | 40 | 22 (55.0) | 18 (45.0) | ||
| F | 39 | 20 (51.3) | 19 (48.7) | ||
| F +/− M +/+ (n = 22) | 7.9 ± 4.7 | All | 177 | 92 (52.0) | 85 (48.0) |
| M | 86 | 42 (48.8) | 44 (51.2) | ||
| F | 91 | 50 (54.9) | 41 (45.1) |
F, Female; M, male.
Minor perturbations in the ovaries of Dp103+/− mice
We sought to determine whether Dp103+/− animals revealed DP103 haploinsufficiency in steroid-producing tissues that are modulated by SF-1 activity. When compared with wild-type mice, Dp103+/− females had heavier ovaries (Fig. 3A). The weight of the testes or adrenal glands was similar between the two genotypes. Histological examination of ovarian sections revealed that even though the number of follicles and corpora lutea were similar between wild-type and Dp103 heterozygous mice, the number of follicles that were devoid of oocytes was significantly higher in heterozygous mice compared with wild-type mice (Fig. 3, B and C). This was accompanied by an increase in the number of DP103-negative oocytes within follicles (Fig. 3C). Empty follicles or DP103-negative oocytes were found at every stage of maturation including primary and antral follicles. We next examined the estrous cycle of wild-type and Dp103 heterozygous females, monitored daily over a period of 21 d. As expected, mice from both genotypes exhibited an estrous cycle (Fig. 4A), yet Dp103 heterozygotes spent more time in estrus compared with their wild-type littermates (Fig. 4B, 42 vs. 27%, respectively, P < 0.03). Because reduced expression of DP103 might influence SF-1 function and consequently the level of steroidogenic enzymes, we sought to examine tissue expression of several pivotal enzymes, including Cyp11a1 (P450scc), Cyp19a1 (aromatase), Cyp17a1 (17α-hydroxylase), steroid acute regulatory protein (StAR), and Cyp21a1 (21-hydroxylase). We found a 3-fold increase in the expression of ovarian Cyp17a1 transcript in Dp103+/− female mice (Fig. 5A), whereas the expression of testicular or adrenal Cyp17a1 was not affected by the genotype. The differences in mRNA expression of all other enzymes tested in the three organs were insignificant (Fig. 5, B and C).
The weight of steroid-producing tissues and ovarian histology. A, Weights of the ovary, testis, and adrenal in male and female mice obtained from wild-type (WT) animals (n = 20) and Dp103 heterozygous (Het) animals (n = 24–26). Values are mean ± sd. *, P < 0.01. B, DP103 expression in ovarian follicles from Dp103 heterozygotes. Sections of adult ovaries were immunostained for DP103 protein, and the presence or absence of DP103 immunofluorescence in follicles was ascertained. C, The frequency of follicular phenotypes as shown in B, in wild-type mice (n = 15–17, one section per animal) and Dp103 heterozygous mice (n = 9–13, one section per animal). Data are represented as mean ± sd. *, P < 0.05 (t test).
Assessment of the estrous cycle in wild-type and Dp103 heterozygous mice. Determination was performed as described in Materials and Methods by investigators blinded to the mouse genotype. A, Daily assessment of estrus (E), diestrus (D), or metestrus/proestrus (M/P); B, fraction of time in estrus for mice in either genotypic group (n = 4 for each genotype). Data were analyzed by combining the total number of days spent in estrus for each of the two groups (based on the data presented in A and compared using χ2 test, P < 0.03).
Increased expression of Cyp17a1 in the ovary of Dp103+/− mice. Expression of key steroidogenic enzymes was assessed by quantitative RT-PCR using total RNA from ovary (A), testis (B), and female adrenal (C). Results are expressed as fold expression (mean ± sd) in Dp103 heterozygote vs. wild-type mice (with levels in wild-type defined as one) and represent three to five experiments, each performed in duplicate. *, P < 0.05 (t test).
Stress response in Dp103+/− mice
Last, we explored the basal and stress levels of ACTH and corticosterone in Dp103+/− mice and their wild-type littermates. As shown in Fig. 6A, Dp103+/− female mice exhibited lower levels of basal ACTH during evening sampling and after stress, compared with wild-type mice. These differences were not associated with altered corticosterone levels (Fig. 6B). All other measurements were not significantly different between the genotypes of both genders, and all mice exhibited a comparable response to 30 min of restraint stress (Fig. 6, A and B).
ACTH and corticosterone levels in wild-type and Dp103+/− mice. Basal plasma hormone levels were measured 2 h after lights on (AM) and 1 h before lights off (PM) or after 30 min of restraint stress, all performed at the same time every morning as described in Materials and Methods by investigators blinded to the mouse genotype. A, ACTH. The differences between stress vs. morning or evening were significant for males and for females (n = 4; P < 0.05, ANOVA with post hoc Bonferroni test). The difference for females at evening or stress were significant (P < 0.05) between the two Dp103 genotypes. The other differences between the two Dp103 genotypes within each paradigm were insignificant. B, Corticosterone. The differences between stress vs. morning or evening were significant for males, and the differences between morning, evening, and stress were significant for females. The differences between the two Dp103 genotypes within each paradigm were insignificant.
Discussion
Our data reveal an essential role for murine DP103 during early embryonic development. We found that Dp103-null mice failed to develop beyond the two-cell stage. Although we did not uncover the cause of the early embryonic lethality, we noted that Dp103-null embryos could not enter a developmental phase that is associated with zygotic gene activation (ZGA), characterized by degradation of maternally inherited transcripts and transition to zygotic gene expression (39–43). Consistent with this observation, DP103 is among the proteins that are selectively up-regulated during ZGA in the two-cell embryo (44, 45). Our findings that 1) haploid oocytes can develop even in the absence of DP103 (albeit at a lower rate compared with wild-type oocytes) and 2) Dp103-deficient, haploid oocytes were able to contribute to Dp103−/− zygotes suggest that germline-derived DP103 protein is not essential before ZGA and raises the speculation that DP103 is involved in the reprogramming that takes places during ZGA and enables embryonic development beyond the two-cell stage.
DP103 harbors domains that are characteristic of DEAD-box RNA helicases. It also exhibits RNA/RNA and DNA/RNA helicase activity (19) and associates with proteins involved in RNA processing, such as Ago2 (29, 46–48) and the SMN protein complex (25, 26, 49–51). The SMN-related complex is required for the assembly of spliceosomal small nuclear ribonucleoproteins (snRNPs) (52). Genetic ablation of murine Smn results in early embryonic lethality before entering the blastocyst stage (49). Similarly, genetic ablation of gemin2, another component of the SMN complex, results in survival beyond the blastocyst stage but preimplantation death (50). Together, these data suggest that the function of DP103 during early zygote development involves RNA processing and conceivably silencing of maternal transcripts.
Unlike Dp103-null mice, heterozygous mice are viable, and although they exhibit several endocrine perturbations, these mice appear healthy and fertile. The relatively minor phenotype associated with Dp103 heterozygous mice is consistent with the observation that at least 70% of small interfering RNA-dependent depletion of DP103 protein is necessary to uncover a measurable effect on assembly of the core SMN complex (51, 53, 54). We focused our analysis on steroid-producing organs of Dp103 heterozygous mice because of previous findings by our group and others regarding the repression of SF-1 activity by DP103 (19, 20). The role of this repression in the reproductive physiology is unknown. We found several endocrine alterations that were limited to the Dp103 heterozygous female. These included larger ovaries, a greater number of empty follicles or DP103-negative oocytes, a prolonged estrous phase, up-regulation of ovarian Cyp17a1 mRNA, and lower levels of basal nocturnal ACTH. Because Cyp17a1 regulates the synthesis of androstenedione, a steroid with androgenic properties (55), our data suggest that the increased occurrence of empty follicles in Dp103+/− mice may result from the perturbation of the delicate steroidogenic environment required for proper follicular development (56). Cyp17a1 is coexpressed with Cyp19a1 during the murine estrous cycle (57, 58). The observation that the expression of Cyp19a1, as well as that of Cyp11a1 and steroid acute regulatory protein, is not enhanced in Dp103 heterozygotes suggests that the increase in Cyp17a1 mRNA is specific, and does not reflect a difference in the estrous cycle stage. Moreover, these observations suggest that the increase in size of the ovary in Dp103 heterozygotes does not reflect the proliferation of one of the enzyme-expressing cell types (59), which would have resulted in enhanced expression of other enzymes. The lower evening ACTH level in the face of similar corticosterone concentrations suggests altered induction kinetics of the circadian corticosterone peak with accentuated feedback inhibition on pituitary corticotrophs or hypothalamic neurons.
Our findings highlight overlapping and distinct molecular pathways regulated by SF-1 and DP103. 1) Dp103-null mice die well before the initial expression of SF-1, which occurs at mouse embryonic d 9 (36, 60). 2) Although both DP103 and SF-1 are highly expressed in the gonads and adrenal glands, their expression pattern revealed marked differences. Whereas ovarian DP103 and SF-1 are both expressed in follicular cells and the corpus luteum, DP103 is highly expressed in oocytes, which do not express SF-1 (34, 35, 38). In the testis, DP103 protein is highly expressed in germ cells, whereas SF-1 is predominantly expressed in Leydig and Sertoli cells (36). In the adrenal gland, both proteins are expressed in the cortex, but DP103 is also expressed in the medulla. 3) The level of expression of several of the steroidogenic enzymes that are regulated by SF-1 was unchanged by DP103 haploinsufficiency. Interestingly, we previously showed that overexpression of DP103 attenuated SF-1-induced Cyp11a1 promoter activity in vitro (19), yet the mRNA level of this rate-limiting steroidogenic enzyme was unchanged in Dp103 heterozygote mice. Together, our data do not illuminate a discrete pathway in which DP103 might negatively regulate the transcriptional activity of SF-1. We also do not know whether DP103 haploinsufficiency during gonad development might stimulate SF-1-dependent ovarian cell proliferation and growth, as has been suggested in adrenocortical cells that overexpress SF-1 (61). It is also possible that the level of DP103 in heterozygous mice is sufficient for repression of SF-1 or that a compensatory mechanism for DP103 haploinsufficiency is up-regulated. Last, we cannot rule out the possibility that Dp103+/− mice could perform differently from wild-type mice under conditions that challenge reproductive performance.
Acknowledgments
We thank Elena Sadovsky for technical support and Lori Rideout for assistance during preparation of the manuscript.
The project described was supported by Grant R01 HD37571 (to Y.S.) from the National Institute of Child Health and Human Development. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Child Health and Human Development or the National Institutes of Health.
Disclosure Statement: J.M., X.Y., Q.O., L.J., P.A., and Y.S. have nothing to declare. L.J.M. consults as a grant reviewer for the Burroughs Wellcome Fund.
Abbreviations
- CEC
Cornified epithelial cells
- DP103
DEAD-box protein-103
- EGFP
enhanced green fluorescent protein
- ES
embryonic stem
- NEC
nucleated epithelial cells
- PMN
polymorphonuclear cells
- SF-1
steroidogenic factor-1, SMN, survival motor neuron
- ZGA
zygotic gene activation
Author notes
J.-F.M. and X.Y. contributed equally to this work.
