The transcription factors SIX3 and VAX1 are required for suprachiasmatic nucleus circadian output and fertility in female mice

Abstract

The homeodomain transcription factors sine oculis homeobox 3 (Six3) and ventral anterior homeobox 1 (Vax1) are required for brain development. Their expression in specific brain areas is maintained in adulthood, where their functions are poorly understood. To identify the roles of Six3 and Vax1 in neurons, we conditionally deleted each gene using Synapsin^cre , a promoter targeting maturing neurons, and generated Six3^syn and Vax1^syn mice. Six3^syn and Vax1^syn females, but not males, had reduced fertility, due to impairment of the luteinizing hormone (LH) surge driving ovulation. In nocturnal rodents, the LH surge requires a precise timing signal from the brain's circadian pacemaker, the suprachiasmatic nucleus (SCN), near the time of activity onset. Indeed, both Six3^syn and Vax1^syn females had impaired rhythmic SCN output, which was associated with weakened Period 2 molecular clock function in both Six3^syn and Vax1^syn mice. These impairments were associated with a reduction of the SCN neuropeptide vasoactive intestinal peptide in Vax1^syn mice and a modest weakening of SCN timekeeping function in both Six3^syn and Vax1^syn mice. Changes in SCN function were associated with mistimed peak PER2::LUC expression in the SCN and pituitary in both Six3^syn and Vax1^syn females. Interestingly, Six3^syn ovaries presented reduced sensitivity to LH, causing reduced ovulation during superovulation. In conclusion, we have identified novel roles of the homeodomain transcription factors SIX3 and VAX1 in neurons, where they are required for proper molecular circadian clock function, SCN rhythmic output, and female fertility.

Keywords: PER2::luciferase; circadian; conditional knock-out; luteinizing hormone surge; six homeobox 3; ventral anterior homeobox 1.

FIGURE 1

SIX3 and VAX1 expression patterns and overlap with Synapsin^cre. (a) Allen Brain Atlas sagittal in situ hybridization images of Six3 and Vax1 expression patterns in the brain at postnatal day 4 (P4) and P56 in male. Scale bars 500 μm. (b) Synapsin^cre efficiently recombines the Rosa26 tdTomato allele in the SCN in Synapsin^cre:tdTomato mice. (i) tdTomato (magenta) presence in the adult female SCN. Scale bar 100 μm. (ii) Sagittal section containing the SCN, stained for tdTomato (RFP). Scale bar 100 μm. (iii) sagittal image of tdTomato (RFP) in an adult male (scale bar 500 μm). Dashed lines indicate zoomed area in (ii). (c) Synapsin^cre recombines the Rosa26 tdTomato allele in the pituitary (female), ovary, and uterus, and not in the testis. Scale bar 500 μm. (d) PCR to assay Synapsin^cre driven recombination of the Six3^flox and Vax1^flox alleles in the female brain. Acc, nucleus accumbens; BLA, basolateral amygdala; CA2, hippocampal area CA2; CG1/2, cingulate cortex; Cpu, caudate putamen (striatum); DG, dentate gyrus; MCPO, magnocellular preoptic nucleus; MEA, medial amygdala; Olf, olfactory bulb; Pir, piriform cortex; PVN, paraventricular nucleus; SCN, suprachiasmatic nucleus; SON, supraoptic nucleus; VL, ventrolateral thalamic nucleus; VP, ventral pallidum. Representative images of n = 3

FIGURE 2

Disrupted LH surge in Six3^syn and Vax1^syn female mice causing female subfertility. (a,b) A 4-month fertility study evaluating the number of litters born. (a) Synapsin^cre male x B6 female (Syn^cre), Six3^fl/fl x Six3^fl/fl (Six3^fl/fl) matings and Six3^syn x Six3^fl/fl or (b) B6 female x Synapsin^cre male (Syn^cre), Vax1^fl/fl x Vax1^fl/fl (Vax1^fl/fl) and Vax1^syn x Vax1^fl/fl matings. Statistical analysis by one-way ANOVA, compared to control (Syn^cre); ** p < 0.01; *** p < 0.001, n = 5–11. (c,d) Estrus cycling was monitored daily in adult females. DI/II, diestrus I/II; E, estrus; P, proestrus; (e,f) % time spent in each cycle. Statistical analysis by two-way ANOVA, n = 6–8. (g,h) Representative H&E images of ovarian morphology. CL, corpora lutea counted throughout the ovary, t test, p > 0.05, n = 4–5. Magnification ×4. (i) Synapsin^cre efficiently recombines the Rosa26 tdTomato (magenta) allele in female neurons, but does not appreciably target GnRH neurons (green) in the preoptic area, scale bar 100 μm. Representative images of n = 4. (j) Illustrative images of GnRH immunohistochemistry in the preoptic area of a female mouse in adulthood. Magnification ×20. (k) Quantification of number of GnRH neurons in male and female Vax1^fl/fl and Vax1^syn mice as determined by immunohistochemistry for GnRH, n = 3–5, Student’s t test (for total numbers of GnRH neurons) and two-way ANOVA for comparisons across Bregma *p < 0.05; **p < 0.01; ***p < 0.001. (I) qRT-PCR of whole hypothalamus for Gnrh1 and Gapdh in Six3^fl/fl and Six3^syn males. Student’s t test, p > 0.05, n = 5–6. (m) LH release in response to a bolus GnRH challenge (10 min) in diestrus/metestrus females. Student’s t test, p > 0.05. n = 4–8. (n) LH release in response to a bolus kisspeptin challenge (Six3^syn 10 min, Vax1^syn 15 min) in diestrus/metestrus females. Student’s t test, *p < 0.05, n = 4–7. (o) LH levels were evaluated in ovariectomized, estrogen primed females at ZT4 and ZT12.30, and fold change of ZT12.30/ZT4 in LH levels determined. Student’s t test, *p < 0.05, n = 3–7

FIGURE 3

Disrupted free-running locomotor activity rhythms in Six3^syn females. (a) Running-wheel activity patterns in Six3^fl/fl (n = 8) and Six3^syn (n = 13) female mice. Data show double plotted actogram activity with 10 days in LD12:12 (LD) followed by 28 days in constant darkness (DD). Data are presented in ClockLab normalized format. Horizontal bar above the actograms indicates lights on (white) and lights off (black) during the LD12:12 cycle. (b–d) Female 14-day average data for indicated analysis parameter in LD12:12 and DD. (e) χ² periodograms for females during 2 weeks in DD. Significance line (set at 0.001) is depicted in red. (f) Running-wheel activity profiles adjusted for estimated free running period in Six3^fl/fl and Six3^syn female mice in DD. SEM is indicated by orange shading. Estimated tau calculated via χ² were as follows; a1 tau = 23.3, a2 tau = 23.5, a3 tau = 24, a4 tau = 24.2. Statistical analysis by two-way repeated measures ANOVA mixed effect analysis, n = 7–13. Matching codes (a1, a2, etc.) on the upper right corner of each actogram, χ² periodogram, and activity profile indicate data from a particular mouse

FIGURE 4

Disrupted wheel-running activity in DD of Vax1^syn females. (a) Running-wheel activity patterns in Vax1^fl/fl (n = 7) and Vax1^syn (n = 10) female mice. Format and indications as in Figure 3. (b–d) Female 14-day average data for indicated analysis parameter in LD12:12 and DD. (e) χ² periodograms for females during 2 weeks in DD. Significance line (set at 0.001) is depicted in red. (f) Running-wheel activity profiles adjusted for estimated free running period in Vax1^fl/fl and Vax1^syn female mice in DD. SEM is indicated by blue shading. Estimated tau, calculated via χ² were as follows; a1 tau = 23.5, a2 tau = 23.5, a3 tau = 23.6, a4 tau = 23.8. Statistical analysis by two-way repeated measures ANOVA mixed effect analysis, n = 7–13. Matching codes (a1, a2, etc.) on the upper right corner of each actogram, χ² periodogram and activity profile indicate data from a particular mouse, p < 0.05

FIGURE 5

VAX1, but not SIX3, alters VIP and AVP neuropeptide expression in the SCN. Illustrative coronal images and quantification of (a–d) H&E, and immunohistochemistry of (e–h) VIP and (i–l), AVP in adult SCN of Syn^cre, Six3^fl/fl, Vax1^fl/fl, Six3^syn, and Vax1^syn mice. Scale bar 100 μm. All H&E and immunohistochemistry experiments were done in three to six male and female brains per genotype. (b,d,f,h,j,l) Data are represented as average ± SEM. T test or one-way ANOVA. *, p < 0.05. Transient transfection of NIH3T3 cells with (m) Vip-luciferase plasmid reporter and (n) Avp-luciferase with or without Six3 overexpression vector, Vax1 overexpression vector, or empty vector (EV) plasmids. Data are expressed as fold change from EV. N = 4–8 independent experiments in duplicate or triplicate. One-way ANOVA as compared to EV; ***p < 0.001

FIGURE 6

SIX3 and VAX1 regulate Per2 expression in NIH3T3 cells and circadian period in SCN explants. (a,b) Transient transfections of NIH3T3 cells with the mouse Per2 regulatory region driving luciferase (Per2-luciferase) with and without Six3 overexpression vector (200 ng) or its empty vector (EV, psg5 200 ng), or Vax1 overexpression vector (20 ng) or its empty vector (EV, pCMV6 20 ng). P is the promoter, E1 is the first exon, and I1 is the first intron for the mouse Per2 gene. Numbers indicated with the stars on the regulatory regions refer to ATTA sites that have been mutated (see Table 1). Statistical analysis by two-way ANOVA mixed effect model, *p < 0.05; **p < 0.01; ***p < 0.001, n = 4–8 in triplicate. (c–f) PER2::LUC recordings in a LumiCycle of ex vivo SCN obtained from (c), Six3^syn:PER2::LUC and (d), Vax1^syn:PER2::LUC diestrus/metestrus females. (e) Histogram of PER2::LUC circadian period of SCN from diestrus/metestrus females. (f) Histogram of PER2::LUC amplitude of SCN from diestrus/metestrus females. Statistical analysis by two-way ANOVA mixed effect model, ***p < 0.001, n = 4–8

FIGURE 7

Mistimed PER2::LUC expression in the SCN and pituitary is associated with reduced ovulatory efficiency in Six3^syn females. (a,b) PER2::LUC phase (time of first peak) in the SCN, pituitary, and ovary of Six3^syn:PER::LUC and Vax1^syn:PER2::LUC diestrus/metestrus females. Data were analyzed by circular analysis of variance high concentration F-test, where * indicates significantly different phases, n = 3–19. (c) Oocytes collected in the fallopian tubes after superovulation. Student’s t test, *p < 0.05, n = 5. (d) Average litter size of the first three litters generated by Six3^syn females paired with Six3^fl/fl males and Vax1^syn females paired with Vax1^fl/fl males for 4 months. Statistical analysis by one-way ANOVA, n = 4–8. (e) Quantification of % follicles with positive LH receptor (LHR) immunohistochemistry staining in ovaries collected at ZT2 and ZT12 in diestrus/metestrus Vax1^fl/fl and Vax1^syn females (n = 3–6), and (f), corresponding illustrative images at ZT12. Black arrow indicates positive staining in theca cells, scale bar 500 μm