Estrogen Receptor-β Action in the Ovary, but Not in the Hypothalamic-Pituitary Axis, is Required for the Generation of Normal Luteinizing Hormone Surges in Mice.

Estrogen receptor-β (ERβ) is expressed in the hypothalamic-pituitary (HP) axis and gonads. ERβ-null (βERKO) female mice display defects in ovarian function and are subfertile, as characterized by fewer pregnancies and smaller litters relative to wild type (WT). We have also shown that the mean amplitude of the proestrus luteinizing hormone (LH) surge is reduced in βERKO female mice compared to WT (6 vs. 22 ng/ml). We hypothesized that the submaximal LH surges in βERKO females are due to functional defects in the ovary that result in less than sufficient estradiol (E2) production, leading to improper stimulation of the HP axis, rather than being due to defects within the HP axis itself. Ovaries from 21 day old βERKO mice were transplanted bilaterally into ovariectomized WT females (βW; n = 5) and vice versa (Wβ; n = 5); WT to WT control transplants (WW; n = 4) and mice not undergoing surgery were also included in the study (WT; n = 5; βERKO; n = 12). Seven weeks after surgery, mice were housed in close proximity to adult male mice, in a 10 h-D 14 h-L lightcycle (D = 2100 h) and estrous cycle stages were determined by daily vaginal cytology for 5 additional weeks. Serial proestrus blood samples were taken from each mouse by mandibular puncture at 1900 h, 2100 h, and 2300 h. Serum LH levels were measured using a Dissociation Enhanced Lanthanide Fluoro-Immuno-Assay (DELFIA) (lower detection level 0.03 ng/ml). Only the peak LH value from each proestrus bleed was used in the analysis. Consistent with our last study, LH surges were defined as LH > 0.43 ng/ml and ranged from 0.5 - 42 ng/ml in WT, 5 - 47 ng/ml in WW, 0.5 - 51 ng/ml in Wβ, 1 - 9 ng/ml in βW, and 0.4 - 42 ng/ml in βERKO. Means ± SE were 28 ± 11 ng/ml in WT, 28 ± 7 ng/ml in WW, 26 ± 7 ng/ml in Wβ, 7 ± 1 ng/ml in βW, and 8 ± 3 ng/ml in βERKO. These data show that the average amplitude of the LH surge was dependent on the presence of ERβ in the ovary, and was not effected by the presence or absence of ERβ in the HP axis. Based on our hypothesis that βERKO ovaries may be unable to produce the estrogen milieu needed for the LH surge to occur, we then supplemented a second group of WT and βERKO mice with 10 µg/kg E2 (n = 7 per genotype) or sesame oil (n = 7 per genotype) on the afternoon of diestrus, and collected blood samples on the following evening (proestrus) to assess LH levels. Peak LH ranged from 0.6 - 35 ng/ml in WT-oil, and 0.8 - 64 ng/ml in WT-E2; means ± SE were 13 ± 6 ng/ml in WT-oil, and 24 ± 9 ng/ml in WT-E2; medians were 8 ng/ml in WT-oil, and 25 ng/ml in WT-E2. βERKO mice were not affected by the E2 treatment (mean peak LH; oil: 4 ± 2 ng/ml vs. E2: 3 ± 1 ng/ml) suggesting that additional ovarian factors besides preovulatory E2 may be altered in βERKO mice. These data further support the need for ERβ in the ovary and less so in the HP axis in regards to the generation of the LH surge. Supported by NIEHS Intramural Research Program.