Female fertility does not require Bmal1 in suprachiasmatic nucleus neurons expressing arginine vasopressin, vasoactive intestinal peptide, or neuromedin-S

Abstract

Disruptions to the circadian system alter reproductive capacity, particularly in females. Mice lacking the core circadian clock gene, Bmal1, are infertile and have evidence of neuroendocrine disruption including the absence of the preovulatory luteinizing hormone (LH) surge and enhanced responsiveness to exogenous kisspeptin. Here, we explore the role of Bmal1 in suprachiasmatic nucleus (SCN) neuron populations known to project to the neuroendocrine axis. We generated four mouse lines using Cre/Lox technology to create conditional deletion of Bmal1 in arginine vasopressin (Bmal1^fl/fl:Avp^cre ), vasoactive intestinal peptide (Bmal1^fl/fl:Vip^cre ), both (Bmal1^fl/fl:Avp^cre+Vip^cre ), and neuromedin-s (Bmal1^fl/fl:Nms^cre ) neurons. We demonstrate that the loss of Bmal1 in these populations has substantial effects on home-cage circadian activity and temperature rhythms. Despite this, we found that female mice from these lines demonstrated normal estrus cycles, fecundity, kisspeptin responsiveness, and inducible LH surge. We found no evidence of reproductive disruption in constant darkness. Overall, our results indicate that while conditional Bmal1 knockout in AVP, VIP, or NMS neurons is sufficient to disrupted locomotor activity, this disruption is insufficient to recapitulate the neuroendocrine reproductive effects of the whole-body Bmal1 knockout.

Keywords: arginine vasopressin; circadian clock; fertility; luteinizing hormone surge; neuromedin-s; suprachiasmatic nucleus; vasoactive intestinal peptide.

Figure 1

Conditional deletion of Bmal1 disrupts home cage activity and temperature rhythms. (A) Representative double-plotted, normalized actograms of male Bmal1^{fl/ fl}, Bmal1^fl/fl:Avp^cre , Bmal1^fl/fl:Vip^cre , Bmal1^fl/fl:Avp^cre+Vip^cre , Bmal1^fl/fl:Nms^cre mice over a 16-week paradigm. Weeks 1-12 were spent in 12:12 LD, with 8-hour phase delays and advances occurring at week 4 and week 8 (arrows). At Week 12, animals were exposed to a 30 min light pulse (*) at ZT 20 and released into constant darkness for the remainder of the experiment (vertical black bar). Data during cage changes have been removed (gray boxes). (B) Body temperature (Tb) rhythms from a representative animal over 10 days in LD (left) and DD (right) compared to a representative Bmal1^fl/fl (gray); data are plotted as mean ± SEM. (C, D) Free-running period (tau) from Chi-squared periodogram over the final 10 days of activity in the first LD period (LD) and constant darkness (DD) (C) or temperature (D) (n = 3-12). (E, F) Chi-squared periodogram amplitude from the same data as C and D for activity (E) and temperature (F). Data for tau and amplitude were analyzed by 2-way ANOVA and Dunnett’s multiple comparison test. (G) Total activity counts over 10 days in LD vs DD for all mutants (n = 3-12). Data were analyzed by 2-way ANOVA and Sidak’s multiple comparisons test. (H) Nocturnality was calculated as the percentage of the activity that occurred during the dark period compared to the total activity during the day for all mutants (n = 3-14). Data were analyzed by one-way ANOVA and Dunnett’s multiple comparison test. (*) p ≤ 0.05; (**) p ≤ 0.01; (***) p ≤ 0.001; (****) p ≤ 0.0001.

Figure 2

Loss of Bmal1 produces modest effects on estrous cyclicity in Bmal1^fl/fl:Nms^cre , but not other mutants. (A-D) Representative cycles over 24 days from Bmal1^fl/fl:Avp^cre , Bmal1^fl/fl:Vip^cre , Bmal1^fl/fl:Avp^cre+Vip^cre , Bmal1^fl/fl:Nms^cre , with respective cohort Bmal1^{fl/ fl} control mice. Percentage time in each estrous stage, average cycle length for completed estrous cycles, and representative cycling graphs from in (E) Bmal1^fl/fl:Avp^cre (n = 5-8), (F) Bmal1^fl/fl:Vip^cre (n = 6-8) (G) Bmal1^fl/fl:Avp^cre+Vip^cre (n = 3-6), and (H) Bmal1^fl/fl:Nms^cre (n = 4-5), analyzed by 2-way ANOVA and Sidak’s post hoc test. (I–L) Estrous cycle length (estrous to estrous) for each line, not significant by Student’s t-test. Significant values were (p ≤.05) are indicated by an asterisk (*).

Figure 3

Fecundity is unaffected by conditional Bmal1 knockout. Time to first litter (A–D), number of litters in 100 days (E–H), and number of pups per litter (I-L) were unaffected for Bmal1^fl/fl:Avp^cre (n = 4-9), Bmal1^fl/fl:Vip^cre (n = 3-6), Bmal1^fl/fl:Avp^cre+Vip^cre (n = 6), and Bmal1^fl/fl:Nms^cre (n = 5-7) mice. Data were analyzed by unpaired t-test.

Figure 4

Exogenous kisspeptin produces an increase in circulating LH in conditional Bmal1 knockout mice comparable to controls. 2 mg/kg kiss-10 significantly increased LH compared to control 10 min after administration in (A) Bmal1^fl/fl:Avp^cre (n = 5), (B) Bmal1^fl/fl:Vip^cre (n = 3-9), (C) Bmal1^fl/fl:Avp^cre+Vip^cre (n = 4-5), and (D) Bmal1^fl/fl:Nms^cre (n = 3-4) mice. Data were analyzed by two-way ANOVA and Sidak’s multiple comparison test. (*) p ≤ 0.05; (**) p ≤ 0.01; (***) p ≤ 0.001; (****) p ≤ 0.0001.

Figure 5

Bmal1 mutants demonstrate an appropriately timed LH surge. LH levels at ZT12 for (A) Bmal1^fl/fl:Avp^cre (n = 7-12), (B) Bmal1^fl/fl:Vip^cre (n = 6-15), (C) Bmal1^fl/fl:Avp^cre+Vip^cre (n = 7-9), and (D) Bmal1^fl/fl:Nms^cre (n = 7-8) mice. AM levels from mice across multiple genotypes and cohorts are plotted for comparison (same in each graph); dashed line represents surge cutoff at two standard deviations above the AM values. For all cohorts, ZT12 LH levels were significantly greater than AM levels, and there was no difference between the control and mutant populations (one-way ANOVA, Tukey’s multiple comparison’s test). (E) Percentage of animals above surge cutoff in each genotype (not significant by Fisher’s exact test).

Figure 6

Exposure to constant darkness does not alter ovarian morphology in conditional Bmal1 mutants. Ovarian histology was performed after six weeks in constant darkness conditions for Bmal1^fl/fl (n = 16), Bmal1^fl/fl:Avp^cre (n = 6), Bmal1^fl/fl:Vip^cre (n = 5), Bmal1^fl/fl:Avp^cre+Vip^cre (n = 4), and Bmal1^fl/fl:Nms^cre (n = 4) mice. Corpora lutea (A, B) Graafian follicles in the ovaries of conditional mutants compared to controls. No significant difference was found between genotypes of either structure (one-way ANOVA).