SEARCH

SEARCH BY CITATION

Keywords:

  • Circadian;
  • ERKO;
  • ESR1;
  • estrogen;
  • NERKI;
  • sex differences;
  • wheel-running activity

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

There are sex differences in free-running rhythms, activity level and activity distribution that are attributed, in part, to the action of gonadal hormones. We tested the hypothesis that non-classical estrogenic signaling pathways at estrogen receptor subtype 1 (ESR1) modify the amplitude and phase of activity. We used ESR1 knock-out mice (ERKO) and non-classical estrogen receptor knock-in mice (NERKI). ERKO animals are unable to respond to estrogen at the ESR1 and NERKI animals lack the ability to respond to estrogens via the estrogen response element-mediated pathway, but can still respond via non-classical mechanisms. We compared intact male and female ERKO, NERKI and wildtype (WT) mice with respect to total wheel-running activity, activity distribution across the 24-h day, phase angle of activity onset and free-running period (τ) and the duration of activity in constant conditions. WT females had significantly greater activity than WT males, and this activity was more consolidated to the dark phase of the light:dark cycle. These sex differences were absent in the NERKI and ERKO animals. Among females, NERKI and ERKO animals had greater activity during the light phase than WT counterparts. Additionally, we have identified a novel contribution of non-classical estrogen signaling pathways on the distribution of activity. Our data suggest that total activity is ESR1-dependent and daily activity patterns depend on both classical and non-classical actions of estrogens. These data will aid in identifying the mechanisms underlying sex differences in sleep–wake cycles and the influence of steroid hormones on circadian patterns.

In humans there are sex differences in circadian rhythms and the epidemiology of healthy sleep and sleep disorders. Objective measures of healthy sleep find sex differences in the total amount of sleep, sleep onset latency and sleep efficiency; there are also sex differences in the prevalence of some sleep disorders, including insomnia, restless leg syndrome and sleep apnea, reviewed in (Krishnan & Collop 2006). The intrinsic circadian period of melatonin secretion and body temperature is shorter in women than in men even when the same bedtimes and wake times are maintained; women tend to wake up earlier and have a greater preference for morning activities than men (Duffy et al. 2011). Further, in women, sleep parameters and the amplitude of body temperature rhythms change from the follicular to luteal phases of the menstrual cycle and change in post-menopausal women following hormone replacement therapy. Together, these data suggest that ovarian hormones modify these daily and circadian rhythms (Leibenluft 1993, Parry et al. 1997, Shibui et al. 2000).

Animal studies provide strong evidence that the exposure to steroid hormones during the development and circulating steroid hormones in adults results in sex differences in the expression of biological rhythms, including sleep–wake patterns. In female rodents, total daily wheel-running activity depends on the stage of the reproductive cycle; when estrogens are elevated, there is an increase in both total wheel-running activity and phase advance in the timing of the daily activity onset in hamsters (Morin et al. 1977), degus (Labyak & Lee 1995, Mahoney et al. 2011) and rats (Wollnik & Turek 1988). In rats, the duration of the active period (α) in constant conditions and in a light:dark (LD) cycle varies throughout the estrous cycle with the greatest duration of the active period occurring on the day of estrus (Albers et al. 1981). Ovariectomy results in reduced total daily activity in both rats and mice (Albers 1981, Ogawa et al. 2003). Intact and estradiol-treated ovariectomized female mice have a longer duration of wake time than males (Paul et al., 2006, 2009). Ovariectomy lengthens free-running period (τ) and estradiol replacement shortens free-running period in female rats (Albers 1981), hamsters (Morin et al. 1977) and degus (Labyak & Lee 1995). There are species differences in the expression of circadian rhythms, for example, in the phase response to light pulse (Daan & Pittendrigh 1976), it is equally possible that the effects of steroid hormones on the expression of circadian rhythms are also dependent upon species.

While the role of estradiol regulation of circadian rhythms in females has been researched, its function with respect to rhythms in males is less clear. Testosterone is the primary gonadal hormone in males and is converted to estradiol and dihydrotestosterone (DHT). Gonadectomy of male mice decreases total wheel-running activity, changes the distribution of activity across the day and increases τ (Daan et al. 1975). Testosterone replacement restores activity and decreases free-running period to that of intact animals. Interestingly, DHT replacement also results in a partial restoration of activity (Karatsoreos et al. 2007), indicating that biological rhythms are regulated by both androgenic and estrogenic actions, at least in males.

The actions of estrogens are primarily mediated through two distinct nuclear receptors: estrogen receptor 1 (ESR1, also known as estrogen receptor α) and ESR2 (also known as estrogen receptor β). To exert action via nuclear receptor-mediated transcriptional activation, estrogens bind to nuclear receptors, which undergo a conformational change promoting dimerization. The hormone-receptor dimer then binds to the estrogen response element (ERE) on regulatory regions of target genes to alter gene transcription. In addition to these ‘classical’ nuclear receptor-mediated mechanisms, there are more recently characterized ‘non-classical’ pathways (Hall et al. 2001, Kousteni et al. 2001). These alternative pathways include the action of the hormone-receptor dimer on non-ERE transcription factors, membrane estrogen receptor-initiated action via protein kinases and ligand-independent ER signaling via second messenger pathways (McDevitt et al. 2008).

Previous studies in this and other laboratories have shown that a lack of estrogens, in gonadectomized animals and in transgenic models, influences the expression of circadian rhythms compared with control animals with endogenous estradiol. Aromatase knock-out mice are unable to synthesize estradiol and show increased activity during the light phase, decreased overall activity levels and increased free-running periods relative to wild-type littermates (Brockman et al. 2011). While these data provide evidence for a role of estrogens in the regulation of circadian patterns, the underlying mechanisms of their action remain relatively unknown. Estrogen receptor knock-out models indicate that the total daily activity level is regulated largely via estrogenic action at ESR1, but other circadian parameters have not been examined (Ogawa et al. 2003). To better characterize the interaction of estrogens and daily and circadian rhythmicity, we are using two transgenic mouse models with alterations in estrogen responsiveness at different points in their response pathway: ESR1 knock-out (ERKO) and ‘non-classical’ estrogen receptor knock-in (NERKI) mice. ERKO mice cannot respond to estrogen via ESR1, but retain the ability to respond at ESR2. NERKI mice have a modified ESR1 receptor knocked into their genome that has a mutation in the ERE-binding domain. In NERKI mice, therefore, estradiol action via ERE-mediated transcription is eliminated but estrogens can still produce effects via other transcriptional pathways (Jakacka et al. 2002). NERKI mice maintain intact ESR2 signaling. This transgenic model allows us to explore effects of estrogens on rhythms that are not mediated by ERE-DNA binding via ESR1.

These transgenic mice provide insight into the mechanism of estrogenic organization of circadian rhythms that differs from gonadectomized mice because they retain circulating gonadally produced steroid hormones, yet have specific mutations in signaling pathways. Further, these mice are uniquely suited for discerning how estrogenic actions at ESR1 organize the expression of circadian rhythms across hormone-sensitive periods of development, rather than an acute disruption with ovariectomy/gonadectomy or estrogen antagonist administration. We will test the hypothesis that estrogens modify the expression of daily and circadian rhythms via ESR1 and via ‘non-classical’ estrogen mechanisms by analyzing male and female wildtype (WT), NERKI and ERKO mice. We are interested in identifying the mechanisms underlying the expression of sex differences under normal physiological conditions thus we used gonadally intact animals throughout these studies.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Animal breeding and care

A total of 65 adult (4–7 months) NERKI, ERKO and WT male and female littermates were used for these experiments. ERKO and NERKI mice were obtained from well-established genetic lines. ERKO mice were derived from an inbred colony of ERaL3−/+ mice (Dupont et al. 2000) maintained at Illinois since it was established in 2005 with animals kindly supplied by Dr. Pierre Chambon (IGBMC, France). NERKI mice [B6.129P2(Cg)-Esr1tm1.1Lja] were obtained from a colony established at the University of Illinois with animals kindly supplied by Dr. J. Larry Jameson (University of Pennsylvania, Philadelphia, PA). NERKI mice were originally created as described (Jakacka et al. 2002). The embryonic stem cell donors for NERKI and ERKO animals were HM-1 and 129/SvPas H1, respectively. Both NERKI and ERKO lines have been bred onto a congenic C57BL/6J background and maintained on this background for the current experiments. ERKO mice are maintained with heterozygous breeding pairs because of infertility of female homozygotes. NERKI mice are maintained by mating heterozygous NERKI males with heterozygous ERKO females because female NERKI heterozygotes are infertile. Control animals are WT littermates of both NERKI and ERKO mice. All genotypes were confirmed using PCR.

Circulating hormone concentrations in ERKO, NERKI and WT mice are relatively similar to one another, allowing us to compare between intact animals. Concentrations of follicle stimulating hormone and luteinizing hormone in WT mice do not differ from ERKO (Eddy et al. 1996) and NERKI (Jakacka et al. 2002) mice. Relative to WT mice, circulating concentrations of estradiol are not significantly different in NERKI males or females (Jakacka et al. 2002). In female ERKO mice, circulating estradiol concentrations are increased compared with WT females (Couse et al. 1995) and in ERKO males testosterone levels are slightly elevated relative to WT males (Eddy et al. 1996).

Breeding animals and litters were maintained on Teklad 8626 rodent diet. Prior to activity monitoring adults were given a Teklad 2016 diet, which contains low soy estrogens (isoflavones) in the range of non-detectable to 20 mg/kg, and remained on this diet for the remainder of the studies. They were given food and water ad libitum. Estrous cycles of all adult females were confirmed via vaginal cytology for 2 weeks prior to behavioral experiments. Cycles were not monitored during the experiments to avoid causing arousal or providing a daily timing cue. In other rodent species, females have a visible increase in wheel-running activity that occurs on the day of estrus, we did not detect a relationship between daily wheel-running activity and estrous cycle stage. Additionally, in a separate group of WT females, stage of estrous cycle did not correlate with daily activity level or phase angle of activity onset (unpublished data).

All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of Illinois and were conducted in accordance with the NIH guide for the care and use of laboratory animals.

Determination of daily activity patterns and circadian parameters

Mice were individually housed in plastic cages (28 × 16 × 12 cm) equipped with a metal wheel affixed to the top of the cage. All mice were housed in light-tight chambers equipped with their own light source. Mice were maintained in 12  h light:12 h dark (LD) cycle unless otherwise indicated. The light intensity at the level of the cages ranged from 220–360 lux (average 290 lux). Wheel revolutions were registered by a magnetic switch and recorded in 10-min bins of activity that were recorded using VitalView and visualized by ActiView (Mini Mitter, Bend, OR, USA). A 10-min bin was defined as active if the number of wheel-revolutions was equal to at least 10% of the animal's peak daily activity rate (in wheel revolutions/10-min bin). The following variables were quantified: average daily wheel revolutions, LD ratio of activity, phase angle of activity onset relative to lights-off, duration of free-running period (τ) and length of active phase (α) in constant light (LL) and constant dark (DD) conditions. All animals were used in all three lighting conditions. Animals were tested in the following order: LD, LL and DD. Between each of these conditions animals were allowed to re-entrain to LD conditions for 7–10 days. For all parameters in 12:12 LD entrainment (activity distribution, total activity, LD ratio and phase angle) measures were calculated from at least five consecutive days.

Temporal distribution of daily activity was determined by the number of wheel revolutions recorded in each 10-min bin over a 24-h time period and averaged across 5 days for each animal. These data were then averaged across each group to map the daily activity trends of WT, NERKI and ERKO males and females. The total daily wheel revolutions for each animal was calculated based on five consecutive days in 12:12 LD. These averages were then used to calculate the group averages for each genotype and sex.

The LD ratio represents the activity during the light phase compared with the dark phase. LD ratios were averaged for 5 days for each individual animal, and then these values were averaged to determine the LD ratio for each group. A ratio close to 1.0 represents an animal that is almost exclusively active during lights-on, while a ratio close to 0 represents an exclusively nocturnal animal.

The phase angle of activity onset is measured relative to the time of lights-off. Positive values indicate that the animal began wheel-running before lights-off and negative values indicate that the activity onset began after lights-off. Phase angle was calculated from activity during LD conditions. The onset of the daily activity bout was defined as the first of three consecutive 10-min bins of activity that were not separated by more than two 10-min bins of inactivity before the next recorded active bout. Using the phase of activity onset, we used circular statistics to calculate a mean vector angle and vector length. The vector angle represents the phase of activity and the vector length reflects the degree of consolidation of the activity to the same time.

The free-running period (τ) was recorded in both LL and DD conditions for at least 7 days. Data from the first 3 days in constant conditions were disregarded to avoid transition effects. The τ value was obtained from a periodogram analysis of wheel-running activity using ActiView. The duration of the period of sustained activity during constant conditions (α) was calculated from 7 days of consecutive activity and averaged for each animal. Activity duration was measured from the onset of the daily activity bout to its cessation. Onset of activity was defined as above for phase angle. Activity cessation was defined as the last active bin in a period of consecutive activity of least 40 min that preceded at least 2 h with no sustained activity. The difference between the τ in DD and the τ in LL for each individual animal was determined (Δτ). Individual Δτ values were averaged for each group.

Statistical analysis

Total daily activity, LD ratio, phase angle of activity onset, τ and α in constant conditions, and Δτ were analyzed with a 2-factor analysis of variance (anova) using sex and genotype as independent factors (systat). A priori post hoc comparisons were performed to determine differences between groups within a sex or within a genotype. The variation in LD ratio and phase angle was quite large, so non-parametric tests were performed to test for effect size (Cohen 1988). This test compares the magnitude of the difference between means independent of the sample size. An effect size of 0.2, 0.5 and 0.8 are small, medium and large, respectively. To statistically compare wheel-running activity during the dark phase, we performed an anova with genotype and time as the independent factors and mean wheel revolutions at 2-h intervals as the dependent variable. We compared males and females separately, and then used a priori post hoc analyses to compare between groups at a given time point. Lights-off was at Zeitgeber time (ZT) 12 (ZT0 = lights-on) and activity was compared at ZT13, ZT15, ZT17, ZT19, ZT21 and ZT23. Differences were significant when P < 0.05.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Sex differences in daily activity patterns are modified by responsiveness to estrogens

Daily activity patterns and total wheel-running activity

The distribution of wheel-running activity across the day depended on sex and genotype. At the time of lights-off, all groups showed an initial surge of activity followed by a decline in wheel-running levels as the dark phase progressed (Fig. 1). To further analyze the temporal pattern of activity, we compared mean activity at 2-h intervals during the dark phase. Within females, there is a significant effect of genotype (F2 = 23.72, P < 0.001) and time (F5 = 10.784, P < 0.001) on activity, but no interaction. All females had increased activity at the beginning of the dark phase, and there were no significant differences in activity level at ZT13 between WT, NERKI and ERKO genotypes. However, as the dark phase progressed, differences in activity levels emerged. ERKO females had significantly less activity than WT females throughout the night (ZT 15, 17, 19, 21, P < 0.05 for all comparisons). NERKI females had intermediate activity levels that were between WT and ERKO animals. They did not differ from ERKO females at any point in the dark phase and were less active than WT females at ZT17 (P < 0.05) and ZT19 (P < 0.05) only. At the end of the dark phase, 1 h prior to lights-on (ZT23), activity levels among all genotypes were the same.

image

Figure 1. Relative patterns of daily activity (a) and representative activity logs (b) for WT, NERKI and ERKO mice grouped by sex. Daily wheel-running activity was measured in 12:12 light:dark cycle (lights-off at ZT 12) for five consecutive days. The number of wheel revolutions was averaged into 1-h bins. Activity levels during the dark phase were compared and differences between groups are indicated: *WT vs. ERKO (P < 0.05), #WT vs. NERKI (P < 0.05) and + NERKI vs. ERKO. Sample sizes are as follows: for female WT (n = 15), NERKI (n = 7) and ERKO (n = 8) and for male WT (n = 7), NERKI (n = 6) and ERKO (n = 19).

Download figure to PowerPoint

Like females, the activity level for males increased dramatically following lights-off (ZT13) then decreased across the night. There was a significant effect of genotype (F2 = 3.721, P = 0.026) and time (F5 = 13.248, P < 0.001) on activity, but no interaction between these variables. Unlike females, there were no significant differences in activity level between WT and either ERKO or NERKI males except at ZT19. At ZT19 ERKO males were less active than WT males (P < 0.05). At ZT23, 1 h prior to lights-on, NERKI males were less active than both WT (P < 0.05) and ERKO males (P < 0.05).

Both WT females and males had peak activity following lights-off, with no significant difference in wheel-running at the beginning of the night at ZT13. However, WT animals exhibited a sex difference in activity level throughout the dark phase. The activity of WT males was significantly reduced compared with WT females at ZT15, ZT17, ZT19 and ZT21 (P < 0.05 at all times). This sex difference in activity levels was not seen in ERKO or NERKI animals at any time point in the dark phase.

We also compared total daily wheel-running activity in all groups. We found a significant effect of genotype on total activity [F2 = 7.365, P = 0.001] but no effect of sex or an interaction between sex and genotype. Among females, WT animals were significantly more active than both NERKI (P < 0.01) and ERKO females (P < 0.01) (Fig. 2). In contrast, among males, there was no significant effect of genotype. We also wanted to identify sex differences within a genotype and post hoc analysis showed WT females were significantly more active than WT males (26 995 ± 2 047 and 18 945 ± 1 689 total wheel revolutions respectively, P < 0.05). This sex difference in total wheel-running activity was abolished in the NERKI and ERKO animals.

image

Figure 2. Average total wheel revolutions ± SEM (a) and average LD ratio ± SEM (b) for male and female WT, NERKI and ERKO mice. Data were averaged over 7 days during a 12:12 light:dark (LD) cycle. WT females had significantly greater total activity than both NERKI (*P < 0.05) and ERKO (*P < 0.05) animals. Among WT mice, females had significantly greater total activity and lower LD ratio than males (a compared with b, P < 0.05). Among males, the LD ratio of WT mice is less than that of ERKO mice (*P < 0.05). Sample sizes as follows: for female WT (n = 15), NERKI (n = 7) and ERKO (n = 8) and male WT (n = 7), NERKI (n = 6) and ERKO (n = 19).

Download figure to PowerPoint

The ability to respond to estrogen influences daily patterns of activity in LD

LD ratio depends on genotype

To compare the distribution of activity across the day we looked at the LD ratio, which describes the amount of activity during the light phase relative to the dark phase of the LD cycle. A greater LD ratio reflects more daytime activity. An anova test identified a significant effect of genotype on LD ratio (F2 = 3.422, P = 0.04) (Fig. 2). There was no effect of sex or an interaction between sex and genotype on LD ratio. WT males were significantly more nocturnal (0.035 ± 0.011) than ERKO males (0.12 ± 0.022, P < 0.05). The NERKI genotype somewhat restores the distribution of activity to the dark phase as the LD ratio of NERKI males (0.075 ± 0.026) is intermediate to WT and ERKO males. The activity patterns in females differed from males, as we found no significant effect of genotype on LD ratio using parametric statistics. However, effect size analysis, which compares the magnitude of the difference independent of sample size, indicates that these groups have a large effect size between WT and both NERKI (d = 0.85) and ERKO (d = 0.82) females.

Similar to our results for daily wheel-running amount, we found a sex difference in LD ratio in WT animals that was absent in NERKI and ERKO genotypes. WT females (LD ratio = 0.014 ± 0.004) were significantly more nocturnal than WT males (LD ratio = 0.035 ± 0.011, P < 0.05).

Phase angle of activity is affected by genotype

The time when animals begin their period of daily activity relative to the time of lights-off is described as the phase angle. Similar to LD ratio and total activity, we found a significant influence of genotype on phase angle (F2 = 5.183, P = 0.009) but no influence of sex or interaction between sex and genotype. Within males, post hoc hypothesis testing indicated that WT animals had an activity onset closer to lights-off than that of ERKO mice (−9 ± 1.3 min vs. −17 ± 2.6 min after lights-off; P = 0.011). NERKI males begin their daily activity relatively later than WT and ERKO males (−27 ± 6.9 min after lights-off), but there are no statistically significant differences between NERKI and either WT or ERKO males (Fig. 3). Non-parametric effect size analysis indicates that these groups have a large effect size (d = 1.44 for NERKI compared with WT males; d = 0.94 for NERKI compared with ERKO males). Among females, the effect of genotype on phase angle resembled that of their male littermates, with NERKI mice beginning activity later than WT and ERKO mice, however, there are no statistically significant differences between groups. Post hoc analysis did not detect any sex differences within any genotype. When data was normalized using circular statistics, we found no significant effects of either sex or genotype on vector angle (phase angle of activity onset) or vector length (consolidation of activity onset).

image

Figure 3. Average onset of activity relative to lights-off, phase angle (min) ± SEM for animals housed in 12:12 light:dark cycle for 5 days. Time represents minutes after lights-off. WT males begin activity closer to lights-off than ERKO males (*P < 0.05). Sample sizes as follows: for female WT (n = 15), NERKI (n = 7) and ERKO (n = 8) and male WT (n = 7), NERKI (n = 6) and ERKO (n = 19).

Download figure to PowerPoint

Circadian rhythms in constant conditions are sex dependent

Duration of the free-running period, τ, depends on sex

In constant darkness the duration of the free-running period, τ, is significantly affected by sex [anova test, (F1 = 5.955, P = 0.018)]. There was no significant effect of genotype or an interaction between genotype and sex on τ in DD. Post hoc hypothesis testing showed that the free-running period of NERKI males (23.9 ± 0.04 h) was significantly greater than that of WT males (23.7 ± 0.04 h, P = 0.04) and NERKI females (23.7±0.04 h, P = 0.039) (Fig. 4). There were no significant genotype differences among females, nor any sex differences between male and female WT or ERKO mice.

image

Figure 4. Average τ ± SEM for animals housed in constant dark (a) or constant light (b). LL and DD y-axis values differ. Values were determined from 7 days of activity. In DD: WT males had a shorter τ than NERKI males (*P < 0.05). NERKI females had a shorter τ than NERKI males (a compared with b, P < 0.05). In LL: For each genotype, there is a significant sex differences (*, +, # P < 0.05). The change in the free-running period from DD to LL is δτ (c). For each genotype, there is a significant sex differences in the Δτ value (*, +, # P < 0.05). Sample sizes as follows: for female WT (n = 15), NERKI (n = 8) and ERKO (n = 8) and male WT (n = 7), NERKI (n = 6) and ERKO (n = 19).

Download figure to PowerPoint

For all groups, the free-running period for animals housed in constant light was significantly lengthened when compared with τ in DD (P < 0.05 for all comparisons). For all groups, free-running rhythms had a period less than 24 h in DD, and greater than 24 h in constant light (LL). As was seen in DD, in LL the free-running period depended on sex [anova test, (F1 = 2.602, P <0.001)]. We did not find a significant effect of genotype or an interaction between sex and genotype. In contrast to data in DD, in LL, post hoc analyses show significant sex differences in τ between males and females for WT (P < 0.05), NERKI (P < 0.05) and ERKO (P <0.05) animals. In each case, females had a longer free-running period than did males. Within a sex, there was no effect of genotype on the free-running period in LL.

Duration of daily activity in constant conditions: α

We calculated duration of activity, described as α, for animals housed in DD. We found a significant influence of genotype (F2 = 3.721, P = 0.03) (Fig. 5), but no effect of sex or an interaction of sex and genotype on α. Among females, WT mice had a shorter α relative to ERKO mice (P <0.05). In contrast, among males, there were no significant genotype differences. NERKI males tended to have an increased α relative to WT males, but this was not significant (P = 0.08). We did not detect any sex differences within any genotype.

image

Figure 5. Average duration of the active period (α) ± SEM for animals housed in constant dark (DD) conditions (a) or constant light (LL) (b) for 7 days. In DD: ERKO females had an increased duration of activity relative to WT females (*P < 0.05). Sample sizes as follows: for female WT (n = 15), NERKI (n = 8), and ERKO (n = 9) and male WT (n = 7), NERKI (n = 6) and ERKO (n = 12). In LL: ERKO males had an increased duration of activity relative to NERKI males (*P < 0.05). Sample sizes as follows: for female WT (n = 11), NERKI (n = 7) and ERKO (n = 8) and male WT (n = 7), NERKI (n = 5) and ERKO (n = 15).

Download figure to PowerPoint

In LL, an anova test identified a significant influence of sex on α (F1 = 6.53, P = 0.014] but no effect of genotype or an interaction between these variables. Among males, ERKO mice had a lengthened α duration relative to NERKI mice (P = 0.042). There were no significant genotype differences among females. There was a significant sex difference in activity duration between WT males and females (P = 0.013) that was absent in NERKI and ERKO genotypes. The difference between the duration of the free-running period in DD relative to LL for each animal is the Δτ. We found a significant effect of sex (F1,54 = 21.97, P < 0.01), but not genotype on Δτ. In all groups, there is a lengthening of the free-running period in LL relative to DD. Females of all genotypes had significantly greater Δτ values than male counterparts (P < 0.05 for all comparisons, data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

We tested the hypothesis that estradiol modifies daily and circadian rhythms via its actions on ESR1 and non-classical signaling pathways. We examined activity parameters in intact animals, as we wanted to understand the mechanisms underlying sex differences in rhythms under physiological conditions as well as identify functional consequences of disrupted estrogenic response on activity patterns. We found significant sex differences in total activity, LD ratio and activity levels during the dark phase of the day in WT animals that are abolished in the NERKI and ERKO genotypes. Our data also indicate that impairments in estrogenic signaling pathways result in changes in the distribution of activity during LD and constant conditions. Lastly, data from NERKI animals showed a novel contribution of non-ERE mediated transcription and estrogen-activated second messenger signaling pathways on LD ratio, and activity levels during the dark phase. Our data suggest that the total daily activity is ESR1-dependent and daily activity patterns (distribution of activity across the day, LD ratio and α) are modulated by both classical and non-classical actions at ESR1. These results provide insight to the mechanisms that underlie sex differences in the hormonal regulation of daily and circadian rhythms.

The most striking effects of altered estrogenic signaling were observed in the distribution and amount of daily wheel-running activity. Compared with males, WT females had sustained and elevated activity across the day (Fig. 1), had more daily activity and were more nocturnal (Fig. 2). Ogawa et al. (2003) described total locomotor activity as dependent on ESR1 in both male and female animals, and did not find sex differences in total wheel-running activity in WT mice. In that study gonadectomized animals were treated with oil or estradiol. We find sex differences in total activity that may be present in intact animals only. Both NERKI and ERKO females have decreased wheel-running activity relative to WT females. These data support the hypothesis that total wheel-running activity is regulated through ERE-mediated actions at ESR1 and provides evidence that there is no contribution from non-classical pathways. Both NERKI and ERKO females have increased LD ratios relative to WT animals. However, NERKI males had an intermediate LD ratio relative to WT and ERKO males. This suggests that there is a sex-specific contribution of non-classical signaling pathways on activity patterns.

Among females, there were significant differences in activity levels between ERKO and WT animals throughout the dark phase. In this measure, NERKI females appeared to have intermediate activity as they did not differ from ERKO or WT females at most of the time points examined. This provides evidence that non-classical signaling, independent of ERE-mediated transcription, contributes to activity distribution in females. Identifying how the independent but related estrogen signaling pathways mediate activity patterns and amplitude, and whether there are sex differences in these mechanisms will provide a foundation for understanding sleep disturbances in humans.

We measured the duration of τ in constant dark and light conditions. In other species, there is an effect of estradiol on the free-running rhythm when animals are housed in constant darkness; ovariectomized female hamsters have a lengthening of τ and administration of estradiol shortens this free-running period (Morin et al. 1977). We predicted that ERKO females, which cannot respond to circulating estradiol via ESR1, would be similar to ovariectomized rodents and thus have a longer τ in DD than WT animals. In DD and LL, however, we did not detect any significant differences in τ among females. In DD, we found a sex difference between NERKI males and females and in LL there were sex differences in τ in all genotypes. Additionally, the change in the duration of τ in LL relative to DD was greater in females than males and did not differ among genotypes within a sex despite differences in the ability to respond to estrogens at ESR1. It is possible that the duration of τ and the change in τ under different lighting conditions are not modified by developmental exposure to estradiol. In contrast, circulating hormones, and the ability to respond to the activational effects of these hormones, may have a modulatory role on the free-running period.

Our data indicate that the contributions of non-ERE mediated transcription and estrogenic regulation of second messenger pathways may not always be additive with the classical pathways. For example, in males, we saw an ‘over-compensation’ in the phase angle of activity onset and the period of the free-running rhythm in DD (Fig. 3 and 4). These effects may be due to compensatory mechanisms in the animal as a result of the impaired ESR1 signaling and also indicate that non-classical signaling pathways can modify the expression of circadian rhythms.

The differences in the expression of circadian activity patterns in adulthood may be due, in part, to a role of estrogens in the regulation of response to photic cues. Estrogens may strengthen the coupling between endogenous behavioral rhythms and the external light environment, as we see more pronounced genotype differences in entrainment (LD) than in free-running conditions. In mammals, the suprachiasmatic nucleus (SCN) of the hypothalamus generates circadian rhythms in activity and estrous cyclicity and entrains activity to the LD cues (Stephan & Nunez 1977). In addition to the ability to respond to light, the SCN is also regulated by internal cues, such as hormonal secretion. The influence of gonadal hormones on daily rhythms may be mediated by acting both centrally on the SCN and peripherally on downstream targets. In humans, there are estrogen receptors (ESR1 and ESR2) in the SCN, with significantly greater ESR1 expression in females relative to males (Kruijver & Swaab 2002). The distribution of estrogen receptors in the mouse brain is less clear, although there is evidence for both ESR1 and ESR2 in the SCN with ESR2 being relatively more abundant (Shughrue 1998, Mitra et al. 2003, Vida et al. 2008). The function of these SCN estrogen receptors remains unknown, although it has been hypothesized that they may not play a major role because of their relatively low expression levels (Karatsoreos et al. 2007). It is probable that estrogens modify behavioral patterns by acting on a number of targets including those downstream from the SCN.

We have identified sex differences in daily and circadian activity patterns and characterized how these differences are modified by the ability to respond to estrogens at ESR1. This initial study sets the foundation for future work, which will identify the contribution of organizational and activational hormone effects on the expression of rhythmic patterns. ERKO and NERKI mice are useful in identifying the organizational role of hormones on circadian rhythms, as these animals have had an impaired ability to respond to endogenous estrogen over their lifespan, and they shed light on the estrogenic mediation of sex differences in circadian behaviors. Future studies will determine if there is a functional consequence of estrogen signaling impairment on the ability to entrain to the LD cycle or respond to circadian disruptions.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  • Albers, H.E. (1981) Gonadal hormones organize and modulate the circadian system of the rat. Am J Physiol Regul Integr Comp Physiol 10, R62R66.
  • Albers, H.E., Gerall, A.A. & Axelson, J.F. (1981) Effect of reproductive state on circadian periodicity in the rat. Physiol Behav 26, 2125.
  • Brockman, R., Bunick, D. & Mahoney, M.M. (2011) Estradiol deficiency during development modulates the expression of circadian and daily rhythms in male and female aromatase knockout mice. Horm Behav 60, 439447.
  • Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Lawrence Erlbaum Associates, Hillsdale, NJ.
  • Couse, J.F., Curtis, S.W., Washburn, T.F., Lindzey, J., Golding, T.S., Lubahn, D.B., Smithies, O. & Korach, K.S. (1995) Analysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor gene. Mol Endocrinol 9, 14411454.
  • Daan, S., Damassa, D., Pittendrigh, C.S. & Smith, E.R. (1975) An effect of castration and testosterone replacement on a circadian pacemaker in mice (Mus musculus). Proc Natl Acad Sci U S A 72, 37443747.
  • Daan, S. & Pittendrigh, C.S. (1976) A Functional analysis of circadian pacemakers in nocturnal rodents - II. The variability of phase response curves. J Comp Physiol 106, 253266.
  • Duffy, J.F., Cain, S.W., Chang, A.M., Phillips, A.J.K., Munch, M.Y., Gronfier, C., Wyatt, J.K., Dijk, D.J., Wright, K.P. Jr. & Czeisler, C.A. (2011) Sex difference in the near-24-hour intrinsic period of the human circadian timing system. Proc Natl Acad Sci U S A 108, 1560215608.
  • Dupont, S., Krust, A., Gansmuller, A., Dierich, A., Chambon, P. & Mark, M. (2000) Effect of single and compound knockouts of estrogen receptors α (ERα) and β (ERβ) on mouse reproductive phenotypes. Development 127, 42774291.
  • Eddy, E.M., Washburn, T.F., Bunch, D.O., Goulding, E.H., Gladen, B.C., Lubahn, D.B. & Korach, K.S. (1996) Targeted disruption of the estrogen receptor gene in male mice causes alteration of spermatogenesis and infertility. Endocrinology 137, 47964805.
  • Hall, J.M., Couse, J.F. & Korach, K.S. (2001) The multifaceted mechanisms of estradiol and estrogen receptor signaling. J Biol Chem 276, 3686936872.
  • Jakacka, M., Ito, M., Martinson, F., Ishikawa, T., Lee, E.J. & Jameson, J.L. (2002) An estrogen receptor (ER)α deoxyribonucleic acid-binding domain knock-in mutation provides evidence for nonclassical ER pathway signaling in vivo. Mol Endocrinol 16, 21882201.
  • Karatsoreos, I.N., Wang, A., Sasanian, J. & Silver, R. (2007) A role for androgens in regulating circadian behavior and the suprachiasmatic nucleus. Endocrinology 148, 54875495.
  • Kousteni, S., Bellido, T., Plotkin, L.I., O'Brien, C.A., Bodenner, D.L., Han, L., Han, K., DiGregorio, G.B., Katzenellenbogen, J.A., Katzenellenbogen, B.S., Roberson, P.K., Weinstein, R.S., Jilka, R.L. & Manolagas, S.C. (2001) Nongenotropic, sex-nonspecific signaling through the estrogen or androgen receptors: dissociation from transcriptional activity. Cell 104, 719730.
  • Krishnan, V. & Collop, N.A. (2006) Gender differences in sleep disorders. Curr Opin Pulm Med 12, 383389.
  • Kruijver, F.P.M. & Swaab, D.F. (2002) Sex hormone receptors are present in the human suprachiasmatic nucleus. Neuroendocrinology 75, 296305.
  • Labyak, S.E. & Lee, T.M. (1995) Estrus- and steroid-induced changes in circadian rhythms in a diurnal rodent, Octodon degus. Physiol Behav 58, 573585.
  • Leibenluft, E. (1993) Do gonadal steroids regulate circadian rhythms in humans? J Affect Disord 29, 175181.
  • Mahoney, M.M., Rossi, B.V., Hagenauer, M.H. & Lee, T.M. (2011) Characterization of the estrous cycle in Octodon degus. Biol Reprod 84, 664671.
  • McDevitt, M.A., Glidewell-Kenney, C., Jimenez, M.A., Ahearn, P.C., Weiss, J., Jameson, J.L. & Levine, J.E. (2008) New insights into the classical and non-classical actions of estrogen: evidence from estrogen receptor knock-out and knock-in mice. Mol Cell Endocrinol 290, 2430.
  • Mitra, S.W., Hoskin, E., Yudkovitz, J., Pear, L., Wilkinson, H.A., Hayashi, S., Pfaff, D.W., Ogawa, S., Rohrer, S.P., Schaeffer, J.M., McEwen, B.S. & Alves, S.E. (2003) Immunolocalization of estrogen receptor β in the mouse brain: comparison with estrogen receptor α. Endocrinology 144, 20552067.
  • Morin, L.P., Fitzgerald, K.M. & Zucker, I. (1977) Estradiol shortens the period of hamster circadian rhythms. Science 196, 305307.
  • Ogawa, S., Chan, J., Gustafsson, J.Å., Korach, K.S. & Pfaff, D.W. (2003) Estrogen increases locomotor activity in mice through estrogen receptor α: specificity for the type of activity. Endocrinology 144, 230239.
  • Parry, B.L., Berga, S.L., Mostofi, N., Klauber, M.R. & Resnick, A. (1997) Plasma melatonin circadian rhythms during the menstrual cycle and after light therapy in premenstrual dysphoric disorder and normal control subjects. J Biol Rhythms 12, 4764.
  • Paul, K.N., Dugovic, C., Turek, F.W. & Laposky, A.D. (2006) Diurnal sex differences in the sleep-wake cycle of mice are dependent on gonadal function. Sleep 29, 12111223.
  • Paul, K.N., Laposky, A.D. & Turek, F.W. (2009) Reproductive hormone replacement alters sleep in mice. Neurosci Lett 463, 239243.
  • Shibui, K., Uchiyama, M., Okawa, M., Kudo, Y., Kim, K., Liu, X., Kamei, Y., Hayakawa, T., Akamatsu, T., Ohta, K. & Ishibashi, K. (2000) Diurnal fluctuation of sleep propensity and hormonal secretion across the menstrual cycle. Biol Psychiatry 48, 10621068.
  • Shughrue, P.J. (1998) Estrogen action in the estrogen receptor α-knockout mouse: is this due to ER-β? Mol Psychiatry 3, 299302.
  • Stephan, F.K. & Nunez, A.A. (1977) Elimination of circadian rhythms in drinking, activity, sleep, and temperature by isolation of the suprachiasmatic nuclei. Behav Biol 20, 116.
  • Vida, B., Hrabovszky, E., Kalamatianos, T., Coen, C.W., Liposits, Z. & Kalló, I. (2008) Oestrogen receptor α and β immunoreactive cells in the suprachiasmatic nucleus of mice: distribution, sex differences and regulation by gonadal hormones. J Neuroendocrinol 20, 12701277.
  • Wollnik, F. & Turek, F. (1988) Estrous correlated modulations of circadian and ultradian wheel-running activity rhythms in LEW/Ztm rats. Physiol Behav 43, 389396.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

The authors thank Rachel McMahon and Lauran Wirfs for technical assistance with animal care and genotyping. We also gratefully acknowledge the support of the Billie Field Graduate Fellowship.