|Home | About | Journals | Submit | Contact Us | Français|
This report introduces a gonadotrope-specific cre transgenic mouse capable of ablating floxed genes in mature pituitary gonadotropes. Initial analysis of this transgenic line, Tg(Lhb-cre)1Sac, reveals expression is limited to the pituitary cells that produce luteinizing hormone beta, beginning appropriately at e17.5. Cre activity is detectable by a reporter gene in nearly every LHβ-producing cell, but the remaining hormone producing cell types and other organs exhibit little to no activity. We used the Tg(Lhb-cre)1Sac strain to assess the role Pitx2 in gonadotrope function. The gonadotrope specific Pitx2 knockout mice exhibit normal expression of LHβ, sexual maturation, and fertility, suggesting that Pitx2 is not required for gonadotrope maintenance or for regulated production of gonadotropins.
Gonadotropes are one of five mature hormone-producing cell types of the anterior pituitary gland. Upon release of GnRH from the hypothalamus, gonadotropes secrete heterodimeric hormones composed of a β-subunit of luteinizing hormone (LHβ) and follicle-stimulating hormone (FSHβ), in combination with a common α-subunit (CGA). These glycoprotein hormones act on the gonads to promote sexual maturity and fertility (Kendall et al., 1995). Transgene ablation of gonadotropes results in near-elimination of the gonadotrope population, hypogonadism, infertility, reduced circulating LH, and reduced levels of PRL (Kendall et al., 1991; Seuntjens et al., 1999).
Identifying the factors that regulate gonadotrope differentiation is a critical step towards understanding fertility at the molecular level. Transcription factors and signaling pathways important for gonadotrope specification and function include Gata2, Pitx2, Nr5a1 (SF1), Egr1, Inhibin, ActRII and ActivRI (Burns and Matzuk, 2002; Charles et al., 2006; Charles et al., 2005; Kumar et al., 2003; Suh et al., 2002; Vesper et al., 2005; Yoshikawa et al., 2000). Examining the roles of some individual genes can be difficult because they are important in many physiological systems essential for embryonic development.
To overcome this problem, cre transgenic lines have been developed with activity in the pituitary gland such as Cga-cre, Nr5a1-cre, and a tetracycline inducible Cga-cre. The Cga-cre transgene targets anterior pituitary cells effectively, although it also exhibits activity in muscle (Charles et al., 2006; Cushman et al., 2000; Zhao et al., 2001). Nr5a1-cre drives cre expression in pre-gonadotropes early in development of the anterior pituitary. Nr5a1 is also expressed in the gonads, adrenal cortex, spleen and hypothalamus, making it difficult to distinguish between the loss of gene function in gonadotropes and the other steroidogenic tissues (Bingham et al., 2006). A transgenic line with tetracycline-inducible Cga-cre activity is effective for gonadotrope-specific excision, but the requirement for drug administration is not compatible with all applications (Naik et al., 2006).
Pitx2 is expressed in the developing and adult anterior pituitary gland (Charles et al., 2005; Gage and Camper, 1997; Semina et al., 1996). Analysis of Pitx2 null mice established the role of Pitx2 in the development of many organs, including the pituitary gland, but mutants die at e14.5 due to severe heart defects (Gage et al., 1999a; Kitamura et al., 1999; Lin et al., 1999; Lu et al., 1999). To study the role of Pitx2 in later pituitary development, a hypomorphic allele (Pitx2neo) was generated. Mice homozygous for the hypomorphic allele live until postnatal day 1 (P1) allowing for pituitary cell specification studies, (Gage et al., 1999a; Suh et al., 2002). Pitx2neo/neo pituitaries lack gonadotropes, and have a decrease in somatotropes and thyrotropes. There is little or no detectable expression of important gonadotrope transcription factors Gata2, Egr1, and Nr5a1 in the hypomorphic pituitaries, demonstrating that the dosage of Pitx2 is critical for the differentiation of the gonadotropes (Suh, et. al., 2002).
In adult mouse pituitaries, the majority of Pitx2-positive cells co-express the gonadotropins or thyrotropin, suggesting that Pitx2 also plays a role in the maintenance of these cell types (Charles, et. al., 2005). To test this hypothesis, we have developed a gonadotrope-specific cre transgenic line to ablate Pitx2 from mature pituitary gonadotropes.
The Tg(Lhb-cre)1Sac transgenic construct (Figure 1a) was generated using a 776 bp sequence from the bovine Lhb promoter (Virgin et al., 1985) fused to eGFP (Kaspar et al., 2002); however, we were unable to detect eGFP in transgenic mice. Thus, founder mice were mated to the RosaGFP cre-reporter mice (B6;129-Gt(ROSA)26Sortm2Sho/J), generating Tg(Lhb-cre)1Sac;RosaGFP progeny. Cre activity was estimated from GFP fluorescence in whole pituitaries (Figure 1b). In the highest expressing line, only the anterior lobe had detectable fluorescent signals. Furthermore, the proportion of discrete RosaGFP positive cells in the anterior lobe is consistent with the gonadotrope population size (Figure 1e, h). We assessed cell specificity by examining colocalization of GFP with LHβ using immunohistochemistry (Figure 1f, i). We detected colocalization of GFP with over 80% of LHβ immunoreactive cells and with little or no LHβ negative cells, indicating that the Tg(Lhb-cre)1Sac line is gonadotrope specific in the pituitary (Figure 1g, j).
We utilized a separate cre-reporter strain to analyze expression of the transgene in other pituitary cell types because colocalization with other pituitary hormones requires co-immunohistochemistry with antibodies generated in the same species. Transgenic mice were mated to the Rosa26LacZ cre reporter mouse strain (officially B6;129S4-Gt(ROSA)26Sortm1Sor/J) (Friedrich and Soriano, 1991; Zambrowicz et al., 1997). Progeny of the cross, Tg(Lhb-cre)1Sac;Rosa26LacZ, yielded varying degrees of lacZ expression, although the penetrance of expression in the majority of pituitaries (n=14) was comparable to expectations for gonadotropes (Figure 2a, c). For each pituitary, 3-6 slides were examined for each hormone. Little to no colocalization is detectable between the lacZ-positive cells and GH, ACTH, PRL or TSHβ in most pituitaries (Figure 2d-g). Rarely, in the Tg(Lhb-cre)1Sac;Rosa26LacZ pituitaries, overlapping expression is detectable with a GH- or TSHβ-producing cell, which may represent normal dual-hormone producing cells (Burrows et al., 1999). The penetrance and specificity of cre reporter activity within LHβ positive cells suggests this line will be an effective tool in the deletion of pituitary transcription factors in gonadotropes.
Tg(Lhb-cre)1Sac;Rosa26LacZ mice were used to determine when cre excision begins (Figure 2h-j). At e16.5, very little to no cre activity is detected by X-gal staining (n=3/3). At e17.5, when endogenous LHβ is detectable by immunohistochemistry, cre activity is detectable (n=3/3). X-gal staining is also present at e18.5 (n=5/5), and in the adult pituitary (Figure 2a).
Adult tissues were examined from Tg(Lhb-cre)1Sac;Rosa26LacZ mice (n=4-8) to ascertain the degree of transgene leakiness in non-pituitary tissues (see Figure 3). There was no evidence of transgene activity in the heart, lung, liver, pancreas, spleen, or skeletal muscle (data not shown). Trophoblast-derived placental tissues are negative, and the visceral yolk sac and chorioallantoic plate are positive at e16.5 (n=3/4, data not shown). No transgene activity is detected in the hypothalamus at e17.5 (a). Transgene activity is detected at very low levels in a few cells in the cortex of the brain (b) and the kidney (c), but not in the adrenal gland (d). Faint X-gal staining is detectable in some corpora lutea and follicles of the ovary in Tg(Lhb-cre)1Sac;Rosa26LacZ mice (e) as well as in non-transgenic mice (f), suggesting background β-galactosidase activity. Activity is evident in some seminiferous tubules of the testes in Tg(Lhb-cre)1Sac;Rosa26LacZ mice (g) but not in non-transgenics. Because active Lhβ transcription has been found in both the rat and human testis, it is possible that the lacZ staining in the testis results from appropriate activity of the Lhβ promoter (Berger et al., 1994; Zhang et al., 1995a; Zhang et al., 1995b).
To determine the importance of Pitx2 in gonadotrope maintenance, we crossed the Tg(Lhb-cre)1Sac mice to Pitx2tm2Sac (Pitx2+/-). The resulting Pitx2+/-;Tg(Lhb-cre)1Sac progeny were mated to Pitx2flox/flox mice to delete Pitx2 post-germ cell development, generating gonadotrope-specific Pitx2-deficient offspring (Pitx2flox/-; Tg(Lhb-cre)1Sac). To verify that Pitx2 deletion was effective in the gonadotropes of the knock-out mice, double-immunohistochemistry was used to detect the colocalization of PITX2 and LHβ in adult pituitaries. In the wild type mouse pituitary, nearly all LHβ cells clearly express PITX2 (Figure 4a). As expected, not all PITX2-positive cells express LHβ since PITX2 is also expressed in thyrotropes (Charles et al., 2005). In the knock-out pituitaries, no PITX2 immunoreactivity was detected in the majority of LHβ cells. The remaining PITX2 expression is likely coming from thyrotropes (Figure 4b). Hence, the Tg(Lhb-cre)1Sac line effectively knocked out Pitx2 in gonadotropes.
To assess the effect of gonadotrope-specific Pitx2 deletion on fertility, four Pitx2flox/-;Tg(Lhb-cre)1Sac mice were each mated with four 6-week old C57BL/6J female mice. All the females mated to the Pitx2flox/-;Tg(Lhb-cre)1Sac males presented a copulation plug and carried full-term pregnancies. Four Pitx2flox/-;Tg(Lhb-cre)1Sac female mice were mated with adult C57BL/6J male mice, and all females were able to carry a full term pregnancy, give live births, and care for pups.
A thorough comparison of the physical phenotype of gonadotrope-specific Pitx2 knock-out mice and wild type mice was performed. Gonadotropin deficiency can cause males to exhibit a diminished growth spurt, resulting in adult males and females of equivalent weight (Davey et al., 1999). Average weights of gonadotrope-specific Pitx2 knock-out mice and their normal, cre-negative litter mates (Pitx2flox/+) were recorded for males and females aged 2-12 weeks. There was no evidence of growth differences in the mutant and normal mice, and males grew larger than females (data not shown). Gonadotropin deficiency can cause striking reductions in the weight of the testes and seminal vesicles (Cunha, 1972). The size and weight of the gonads from both sexes of knock-out and wild type mice were similar. Misregulation of gonadotropins can cause precocious puberty or delayed sexual maturation (Cushman et al., 2001; Kendall et al., 1991; Vesper et al., 2006). The vaginal openings in female knock-out and wild type mice occurred at the same time, as did the detection of major urinary proteins (MUPs) in the urine of male mice (data not shown). Taken together, these data suggest that Pitx2 expression in gonadotropes is dispensable for pituitary-gonadal axis function after birth.
It was surprising that deletion of Pitx2 in pituitary gonadotropes had no apparent effect on gonadal development, puberty, or fertility. Many studies have shown that Pitx2 transactivates the genes encoding gonadotropin subunits Cga, Lhb and Fshb (Suszko et al., 2003; Tremblay et al., 2000). It is possible that Pitx2 is dispensable in differentiated gonadotropes because Pitx1 is capable of compensation. Pitx1 and Pitx2 have overlapping functions in the activation of Lhx3 in Rathke's pouch development at e10.5 and in establishing a normally sized pituitary anlage (Charles et al., 2005; Gage et al., 1999a; Suh et al., 2002). Pitx1 and Pitx2 also have overlapping functions in other organs (Gage et al., 1999a, 1999b; Lanctot et al., 1999; Marcil et al., 2003; Suh et al., 2002). Because Pitx1 mutant pituitaries appear minimally affected, and Pitx2 null pituitaries are extremely underdeveloped, Pitx2 is more important than Pitx1 in pituitary development (Suh et al., 2002). Pitx1 might be more critical in adult pituitary function than Pitx2. This prediction could be tested with a floxed allele of Pitx1 and the Tg(Lhb-cre)1Sac line that we report here.
In this study, we characterize a cre transgenic line that is a valuable tool for deleting genes in gonadotropes and demonstrate its effectiveness by using it to assess Pitx2 function in mature gonadotropes.
Tg(Lhb-cre)1Sac mice were generated using a 776 bp sequence from the bovine Lhb promoter. The gonadotrope specificity of the promoter was demonstrated using a herpes simplex virus thymidine kinase reporter gene (Keri et al., 1994; Virgin et al., 1985). We generated a construct for microinjection that uses the bovine LHβ subunit gonadotrope-specific promoter to drive the expression of an enhanced GFP-cre recombinase fusion gene (Kaspar et al., 2002). We engineered a consensus KOZAK sequence (GCCGCCACCATGG) to encourage efficient translation of the transgene mRNA transcript into protein and inserted a poly adenylation signal and intron from the rabbit β-globin gene for efficient nuclear processing of the transcript (Figure 1a). The complete construct was linearized and microinjected into F2 zygotes from (C57BL/6J × SJL) F1 (Hogan, 1994). Live progeny were genotyped via PCR for the presence of cre recombinase in from genomic DNA prepared from a tail biopsy.
Cre mice were maintained by crossing with C57BL/6J mice from The Jackson Laboratory. All mice were maintained at the University of Michigan under the guidelines of the Unit for Laboratory Animal Medicine and the University Committee for Care and Use of Animals. Cre mice were identified by PCR amplification of genomic DNA with primers 5′-gcataaccagtgaaacagcattgctg-3′ and 5′-ggacatgttcagggatcgccaggcg-3′ under the following conditions: 94°C for 3 minutes, followed by 32 cycles of 94°C for 30 seconds, 60°C for 60 seconds and 72°C for 90 seconds, and a final 10 minute extension at 72°C. For transgene analysis, experimental animals carried one allele of the cre transgene and one allele of the reporter gene, while controls were negative for the cre transgene but positive for a reporter gene. B6;129-Gt(ROSA)26Sortm2Sho/J and B6;129S4-Gt(ROSA)26Sortm1Sor/J reporter mice were obtained from The Jackson Laboratory and maintained as homozygotes. Genotyping of B6;129-Gt(ROSA)26Sortm2Sho/J mice was performed by PCR using primers 5′-ggcttaaaggctaacctgatgtg-3′, 5′gcgaagagtttgtcctcaacc-3′ and 5′-ggagcgggagaaatggatatg-3′ under the following conditions: 94°C for 3 minutes, followed by 35 cycles of 94°C for 30 seconds, 64°C for 60 seconds and 72°C for 60 seconds, and a final 10 minute extension at 72°C. The B6;129-Gt(ROSA)26Sortm2Sho/J band is 1,146 bp and wild-type band is 374 bp. Pitx2flox/-;Tg(Lhb-cre)1Sac mice were generated by mating B6-Pitx2+/- mice with Tg(Lhb-cre)1Sac positive mice. The Pitx2+/-;Tg(Lhb-cre)1Sac offspring were mated to B6-Pitx2flox/flox mice, and genotyping was performed as previously described (Gage et al., 1999a).
Adult pituitaries were collected at six weeks of age or later and fixed for 1 hour in 4% formaldehyde in PBS. X-gal staining was performed as previously described (Brinkmeier et al., 1998). After staining, pituitaries were fixed in 4% formaldehyde overnight, rinsed in PBS, dehydrated and embedded in a Citadel 1000 (Thermo Electric, Chesire, England) paraffin embedding machine, and sectioned coronally at 5μm thickness. Immunohistochemistry for the pituitary hormones was performed as previously described (Kendall et al., 1991). For embryos, noon of the day of the vaginal plug is designated as embryonic day 0.5. Embryos and adult organs were dissected, frozen on dry ice and stored at -80°C. Embryos were embedded in OCT (Sakura Finetek Co., Torrance, CA) and cryosectioned at 16μm thickness. After X-gal staining, sections were counterstained for 2 minutes with 1% neutral red stain plus 4% sodium acetate:glacial acetic acid. Sections were dehydrated and mounted with xylene:permount 1:2 (Fisher) mounting media. Whole mount capture of GFP fluorescence of freshly dissected Tg(Lhb-cre)1Sac × B6;129-Gt(ROSA)26Sortm2Sho/J progeny was achieved using a Leica MZFL III stereo/dissecting fluorescent microscope. Pituitaries were fixed for 1 hour in 4% formaldehyde and rinsed in 1× PBS. Immunohistochemistry for GFP and fluorescent LHβ expression was performed on 8-10μm pituitary cryosections using a rabbit anti-GFP Alexa Fluor 488 (Molecular Probes, Eugene, OR) antibody overnight at 4°C, diluted in a blocking solution comprised of 3% normal donkey serum, 1% BSA, and 0.5% Triton-X100 in 1× PBS. Slides were washed 3 times for 5 minutes using 0.5% Triton-X100 in 1× PBS. Guinea pig anti-LHβ antibody (NHPP) was diluted 1:100 in the same block and incubated for 1 hour at room temperature. Washing with 0.5% Triton-X100 in PBS was followed by a one hr incubation with biotinylated anti-guinea pig secondary antibody (Jackson Immunoresearch), 0.5% Triton-X100/PBS washes. Streptavidin-TRITC was added for 1 hour at 1:200 dilution. Rabbit-anti-PITX2 antibody was generated by Dr. Tord Hjalt (Lund University, Sweden) and provided by Dr. Philip Gage (University of Michigan, Ann Arbor). PITX2 antibody was diluted 1:100 in the same block described above and 100 μl was placed on each slide over night at 4°C. Secondary detection was performed as described above using biotinylated anti-rabbit antibody (Vector Laboratories) then washed in PBS/Triton, mounted with fluorescent mounting media and images were captured using a Leica DMRB fluorescent microscope.
We would like to thank the University of Michigan DNA Sequencing Core, Dr. John Nilson for providing the Lhb promoter, Dr. Parlow at the National Hormone and Pituitary Program for providing the pituitary hormone antibodies, Drs. Philip Gage and Donna Martin for the Pitx2flox/flox mice, Drs. Fred Gage and Hoonkyo Suh for supplying the eGFPcre fusion construct, Kailey Owens for help with genotyping the mice, Galina Gavrilina for preparation of transgenic mice, and the University of Michigan Transgenic Animal Model Core.
Author contributions: MAC prepared the transgene construct and identified the best cre line by breeding with the GFP reporter strain as part of his PhD thesis research. AHM used the Tg(Lhb-cre) line to analyze the role of Pitx2 in gonadotropes. MAP analyzed the developmental, tissue-specific, and cell-specific expression of Tg(Lhb-cre) and played the leading role in preparation of the manuscript.