Hepatocyte Ar knockout mice were maintained in our laboratory as previously described (11). Briefly, we crossed an exon 2-floxed Ar (12, 13) female (AR^fl/fl; Cre^-/-) mouse, with a male albumin-Cre ^+/- mouse to produce developmental hepatic Ar knockout mice (LivARKO, AR^fl/fl; albumin (Alb)-Cre^+/-), The Alb-Cre mouse produces liver-specific expression of Cre driven by the albumin promoter in a C57/BL6 background. Litter mates (AR^fl/fl; Alb-Cre^-/-) were referred to as Control (Con) mice. Genotyping primers were designed to detect the presence of Cre or the floxed allele, WT allele, or knockout allele of Ar. Genomic DNA obtained from tail or ear was used for genotyping. Genomic DNA obtained from the liver will amplify a 952-bp amplicon for floxed Ar allele, if the sequence between the LoxP sites is excised, it will amplify a 404-bp amplicon for Ar KO allele. Primer sets were listed in Supplementary Table 1 . Once female mice reached 2-months old, 4 mm-DHT (DHT) or vehicle (Veh) pellets were inserted to the mice under the skin (14–19). Two months after insertion of the pellets, body weight was recorded. All mice were housed in the Johns Hopkins University mouse facility, and all experiments were conducted under a protocol approved by the Johns Hopkins Animal Care and Use Committee.

RNA isolation was performed on collected tissues from Con-veh and LivARKO-veh mice using Trizol (BioRad) as previous described (20). Ar mRNA levels in liver, ovary, uterus, hypothalamus, pituitary, gonadal fat and muscle were measured by quantitative real-time PCR (qRT-PCR) using iQSYBR green reagent according to the manufacturer’s protocol (Bio-Rad). Briefly, 1µg of RNA was reverse transcribed to cDNA using an iScript cDNA kit (Bio-Rad Laboratories). Real-time qPCR was performed to determine the presence and relative expression levels of Ar mRNA in the various tissues. Real-time qPCR was performed in duplicate using SYBR Green Master Mix (Bio-Rad Laboratories) and the CFX Connect qPCR machine (Bio-Rad Laboratories). For Ar primer set (listed in Supplementary Table 1 ), PCR efficiency was determined by measuring a 10-fold serial dilutions of cDNA and reactions having 95% and 105% PCR efficiency were included in subsequent analyses. Relative differences in cDNA concentration between WT and LivARKO mice were then calculated using the comparative threshold cycle (Ct) method. To compare the difference of Ar expression in the same tissue between WT and LivARKO, a △Ct was calculated to normalize for internal control using the equation: Ct (Ar) – Ct (18S). △△Ct was calculated: △Ct (LivARKO) -△Ct (WT). Relative Ar mRNA levels were then calculated using the equation fold difference = 2△△Ct.

Puberty was assessed beginning at 21 days of age by visual inspection of vaginal opening and assessing the age of first estrus (21). Estrous cyclicity was determined, beginning at 8 weeks of age and continuously for 24 days, by assessing vaginal cytology (21). Fertility was assessed by mating 2-3 month old female mice with proven fertile wild type male mice for 90 days, and recording the number of pups and number of litters per female (21, 22). The examiners were blinded to genotypes during all data collection.

Females implanted with DHT (Con-DHT vs LivARKO-DHT) or an empty pellet (Con-veh vs LivARKO-veh) for 15 days were divided into two groups. Group 1 underwent examination of estrous cyclicity by vaginal cytology for 24 days starting at day 15 after DHT treatment. Females in group 2 were mated with fertile males (one female with one male per cage) for 90 days starting at day 15 of DHT treatment. Fertility was examined as described above.

The ovary was dissected from diestrus mice and fixed in 10% formalin phosphate buffer and sectioned to 5 microns thickness in its entirety by Johns Hopkins Medical Laboratories (Histology group). Every 10th section was collected, and ovarian sections were stained with hematoxylin and eosin and examined with a Zeiss microscope.

Blood samples were collected from submandibular vein (17, 21) between 9:00 and 10:00 AM and basal levels of serum LH and FSH were measured. LH and FSH from serum of mice at diestrus were measured by Luminex assay ((MPTMAG-49K, Millipore, Billerica, MA) on a Luminex 200IS platform (Luminex Corporation). The assay detection limit for LH was 0.012 ng/mL and for FSH was 0.061ng/mL. The intra-assay and interassay coefficients of variation (CV) for LH and FSH were between 5% and 9%.

Mice were fasted overnight (16 h) and received intraperitoneal (i.p.) injections of 2 g/kg body weight (BW) glucose. Glucose level was measured at 0, 15, 30, 60, 90 and 120 minutes after glucose injection.

Statistical analyses used were described in each individual figure legend. Some data were analyzed by student t-tests. Some data were assessed by 2-way ANOVA with main effects of DHT treatment and genotype assessed. All analyses were performed using Prism software (GraphPad, Inc.). All results were expressed as means ± SEM. A value of p<0.05 was defined as statistically significant.

A hepatocyte-specific AR knockout mouse was generated using the CRE/lox system (11). In LivARKO mice, Ar mRNA expression in the liver was significantly reduced (90%) compared to control mice littermates, while Ar mRNA expression was not altered in the ovaries, uterus, brain, pituitary, fat or muscle ( Figure 1A ). The PCR product from liver DNA indicated the homozygous floxed-Ar alleles in control mouse and KO alleles in LivARKO mouse ( Figure 1B ). In our observations (11, 15, 17) and others (10, 23), mice with AR^fl/fl, Alb-Cre ^-/- do not exhibit a different reproductive or metabolic phenotype, thus AR^fl/fl mice are commonly used as control mice.

To assess age of pubertal onset in female mice, age of initial vaginal opening and first estrus was determined. Compared to control littermates, LivARKO females experienced vaginal opening (28.3 ± 0.6 vs 28.4 ± 0.9 day of life) and first estrus (37.1 ± 1.0 vs 37.0 ± 1.3 day of life) at similar ages ( Figures 2A, B ). There was also no difference in ovarian estrous cycle duration or pattern ( Figures 2C–E ); LivARKO female mice spent similar amounts of time in proestrus, estrus, and met/diestrus as control mice. During the mating period of 90 days, there was no significant difference between control and LivARKO mice in either the number of litters or number of pups per female ( Figures 2F, G ).

We have previously reported that low dose DHT to control female mice results in reproductive and metabolic abnormalities (11, 15, 17) and that LivARKO mice do not experience the insulin resistance and glucose intolerance present in control mice upon DHT treatment (11, 19). We examined whether female LivARKO mice, while not exhibiting the metabolic derangements, would experience reproductive dysfunction. In both control and LivARKO vehicle treated mice ( Figure 2C–E ), estrous cycling was disrupted (consistent diestrus) upon treatment with DHT ( Figure 3A ). Estrous cycles in the 4 groups of mice were shown in Figure 3B .

Examples (one mouse as representation per group) of fertility profiles are plotted for Con-DHT and LivARKO-DHT mice in Figure 3C . Number of litters and average litter size was compared among experimental groups ((litters: 3.0 ± 0.0 (Con-veh); 3.2 ± 0.2 (LivARKO-veh); 0.3 ± 0.3 (Con-DHT); 0.0 ± 0.0 (LivARKO-DHT); number of pups: 26.3 ± 3.0; 25.6 ± 4.6; 2.5 ± 2.5; 0.0 ± 0.0)) and graphed in Figure 3D, E . Vehicle treated data are displayed as references. The number of litters ( Figure 3D ) and pups per female ( Figure 3E ) of LivARKO-DHT and Con-DHT mice during the mating period was significantly reduced compared to Con-veh and LivARKO-veh mice. In the presence of hyperandrogenemia, LivARKO-DHT mice failed to exhibit impaired glucose tolerance compared to Con-DHT mice ( Figures 3F, G ).

During the two months of DHT treatment, body mass ( Figure 4A ) was not altered among groups. Morphology of ovaries from Con-veh, LivARKO-veh, Con-DHT and LivARKO-DHT mice is shown in Figures 4B–E . Ovarian architecture was similarly altered upon DHT treatment in both control and LivARKO mice ( Figures 4D, E ). The most marked difference was abundance of the CL ((corpora lutea: 2.7 ± 0.8 (Con-veh); 2.3 ± 0.4 (LivARKO-veh); 0.0 ± 0.0 (Con-DHT); 0.0 ± 0.0 (LivARKO-DHT)) which were much less common in the ovaries of the DHT treated mice ( Figures 4D–F ) than in any of the vehicle treated groups ( Figures 4B, C, F ). There was no significant difference in serum levels of LH and FSH among Con-veh, LivARKO-veh, Con-DHT and LivARKO-DHT groups: LH: 0.36 ± 0.07 vs 0.46 ± 0.09 vs 0.32 ± 0.03 vs 0.56 ± 0.10ng/ml; FSH 0.99 ± 0.50 vs 0.72 ± 0.11 vs 0.95 ± 0.18 vs 0.74 ± 0.13ng/ml ( Figures 4G, H ).