Deletion of Otx2 in GnRH neurons results in a mouse model of hypogonadotropic hypogonadism

Abstract

GnRH is the central regulator of reproductive function responding to central nervous system cues to control gonadotropin synthesis and secretion. GnRH neurons originate in the olfactory placode and migrate to the forebrain, in which they are found in a scattered distribution. Congenital idiopathic hypogonadotropic hypogonadism (CIHH) has been associated with mutations or deletions in a number of genes that participate in the development of GnRH neurons and expression of GnRH. Despite the critical role of GnRH in mammalian reproduction, a comprehensive understanding of the developmental factors that are responsible for regulating the establishment of mature GnRH neurons and the expression of GnRH is lacking. orthodenticle homeobox 2 (OTX2), a homeodomain protein required for the formation of the forebrain, has been shown to be expressed in GnRH neurons, up-regulated during GnRH neuronal development, and responsible for increased GnRH promoter activity in GnRH neuronal cell lines. Interestingly, mutations in Otx2 have been associated with human hypogonadotropic hypogonadism, but the mechanism by which Otx2 mutations cause CIHH is unknown. Here we show that deletion of Otx2 in GnRH neurons results in a significant decrease in GnRH neurons in the hypothalamus, a delay in pubertal onset, abnormal estrous cyclicity, and infertility. Taken together, these data provide in vivo evidence that Otx2 is critical for GnRH expression and reproductive competence.

Fig. 1.

Schematics of constructs used to generate GnRH-Otx2KO mice. Transgenic mice expressing Cre recombinase specifically in GnRH neurons (A) were bred with Otx2 floxed mice (B). A frt recombination site (white triangle) remains from the neomycin excision downstream of the loxP sites (black triangles). Primers used in PCR genotyping are labeled F1 and R1. PCR primers used to detect recombination are labeled P1 and R1. C, Genotyping by PCR analysis of the genomic DNA produced a band migrating at 560 bp in the mice bearing a FL allele and a band at 360 bp in WT mice. D, PCR analysis with primers P1 and R1 yields a band of 300 bp after recombination. This band was seen only in the hypothalamus (H) and not kidney (K), heart (M), pituitary (P), or cerebellum (C) of the GnRH-Otx2KO mice.

Fig. 2.

The total number and distribution of GnRH neurons in the adult brain of GnRH-Otx2KO mice. A, Histogram of the spatial distribution of GnRH immunostaining neurons from control and GnRH-Otx2KO (cKO) mice. The most significant reduction of the GnRH immunostaining neurons in the cKO mice occurs at the OVLT. The inset is a schematic representation of a mouse brain, denoting the relative location of the OVLT and the median eminence. Each bar spans 120 μm, with error bars representing the sem. B, GnRH immunostaining neurons from representative control (i) and cKO (ii) coronal sections at the OVLT. Note the decrease in GnRH neuronal cell bodies and the axonal projections in the cKO mice. Magnification, ×50. Scale bar, 200 μm. C, The total number of GnRH immunostaining neurons are significantly reduced in the cKO mice. *, P < 0.05; ***, P < 0.0001.

Fig. 3.

Decrease of GnRH neurons in forebrain regions at e14.5, e15.5, and PN1 in GnRH-Otx2KO mice. A, Histogram of the total number (±sem) of the GnRH immunostaining neurons, the total number of neurons in the olfactory placode, and the total number of neurons in the CNS for e14.5, e15.5, and PN1 control and GnRH-Otx2KO (cKO) mice. There are significantly fewer total GnRH neurons in GnRH-Otx2KO mice at each time point, and although there are similar numbers of GnRH neurons in the nasal placode, there is a significant reduction in neurons in the CNS of cKO mice at each time point. B, Photomicrographs of 40-μm sagittal sections of e14.5 control (i–iii) and GnRH-Otx2KO (iv–vi) mice immunostained for GnRH at different magnifications. In the cKO mice, fewer GnRH neurons migrate across the CP from the OP into the FB region. Magnifications, ×50 (i and iv), ×100 (ii and v), and ×200 (iii and vi). Scale bars, 200 μm (i and iv), 100 μm (ii and v), 50 μm (iii and vi). *, P < 0.05; **, P < 0.001.

Fig. 4.

Total number of EYFP immunostaining GnRH neurons is decreased in the adult brain of GnRH-Otx2KO-EYFP mice. A, Schematic of the mating paradigm to generate control and GnRH-Otx2KO mice expressing EYFP. The PGK-neo selectable marker and tpA transcriptional stop sequence are flanked by loxP sites (black triangles). After the GnRH-Cre-mediated excision of the loxP flanked (PGK-neo, tpA) cassette, the EYFP transgene is specifically expressed in GnRH neurons. B, The total number of EYFP immunostaining neurons is significantly reduced in GnRH-Otx2KO-EYFP mice, indicating a defect in migration/survival rather than expression. **, P < 0.005.

Fig. 5.

GnRH-Otx2KO mice have an increased rate of programmed cell death along the GnRH neuron migratory route. A, Histogram (±sem) showing significantly more cleaved caspase-3 immunostaining in the olfactory placode of GnRH-Otx2KO (cKO) mice than controls at e14.5, with the majority of apoptotic events occurring at the CP. Cell death was similar in the forebrain of both genotypes. *, P < 0.05. B, Photomicrographs of 40-μm sagittal sections of control (i) and GnRH-Otx2cKO (ii) mice immunostained for cleaved caspase-3 at e14.5. Apoptotic cells can be seen along the migratory route from the OP, at the CP, and in the FB region. Magnification, ×200. Scale bar, 50 μm.

Fig. 6.

Hypothalamic GnRH and pituitary LHβ and FSHβ subunit mRNA levels are reduced in GnRH-Otx2KO mice. Quantitative PCR assessment of hypothalamic GnRH (A), pituitary LHβ subunit (B), and FSHβ subunit (C) mRNA levels demonstrate a significant reduction of GnRH and LHβ and FSHβ subunit expression in GnRH-Otx2KO (cKO) mice in both females and males. *, P < 0.05; **, P < 0.01; ***, P < 0.0005.

Fig. 7.

Serum hormone levels in GnRH-Otx2KO animals are also reduced. Although circulating nonsurge LH levels (A) were not significantly different, FSH (B), estradiol (C), and testosterone (D) levels were significantly decreased in female and male GnRH-Otx2KO (cKO) mice. *, P < 0.05; ***, P < 0.0001.

Fig. 8.

Reproduction in GnRH-Otx2KO mice is impaired. Both male and female GnRH-Otx2KO (cKO) mice produced fewer litters over 90 d (A) and had fewer pups per litter (B) as compared with controls. C, Schematic representation of GnRH-Otx2KO (cKO) females' reproductive history. Diamonds indicate new litter, whereas X indicates an infertile female. Five of nine female cKO mice failed to produce any litters over 90 d, and only one female produced more than two litters during this period. *, P < 0.05; ***, P < 0.0005.

Fig. 9.

Cycling profiles show a state of persistent estrus in GnRH-Otx2KO females. A, Profile of individual daily vaginal smears for representative control (i and ii) and GnRH-Otx2KO (cKO) (iii and iv) female mice. Control mice progressed normally from estrus (E), to metestrus/diestrus (D/M) to proestrus (P), whereas cKO mice predominantly remained in estrus. B, The increase in time spent in estrus for cKO female mice was offset by a decrease of time spent in proestrus. Each animal was staged for at least 19 consecutive days. *, P < 0.05; **, P < 0.005.

Fig. 10.

The LH surge is absent in GnRH-Otx2KO females. Profile of LH serum levels (triangles) overlaid on the cycling profile (squares) for representative control (i–ii) and GnRH-Otx2KO (cKO) (iii–v) female mice. The control animals had an LH peak of 31.5 ± 1.0 ng/ml (n = 3) on the evening of proestrus, but cKO animals rarely had a proestrus stage and displayed no change from basal LH levels. The limit of detection in this assay is 48 pg/ml.