Report

Recovery from Age-Related Infertility under

Environmental Light-Dark Cycles Adjusted to the
Intrinsic Circadian Period
Graphical Abstract Authors
Nana N. Takasu, Takahiro J. Nakamura,
Isao T. Tokuda, Takeshi Todo, Gene D.
Block, Wataru Nakamura

Correspondence
gblock@conet.ucla.edu (G.D.B.),
wataru@dent.osaka-u.ac.jp (W.N.)

In Brief
Takasu et al. show that weekly
perturbation of the light-dark cycle
disrupts estrous cycles in middle-aged
female mice, and early infertility evident in
female mice deficient in core circadian
clock genes is improved by coordinating
environmental and endogenous circadian
rhythms. These findings suggest that an
age-related decline in fertility may be
rescued by control of environmental
timing signals.

Highlights
d Environmental perturbation disrupts regular estrous cycles in
middle-aged females

d Cry-deficient female mice show accelerated senescence of

fertility

d Optimal circadian periods rescue female mice from age-

related infertility

Takasu et al., 2015, Cell Reports 12, 1407–1413

September 1, 2015 ª2015 The Authors
http://dx.doi.org/10.1016/j.celrep.2015.07.049
Cell Reports

Report

Recovery from Age-Related Infertility

under Environmental Light-Dark Cycles Adjusted
to the Intrinsic Circadian Period
Nana N. Takasu,1 Takahiro J. Nakamura,2 Isao T. Tokuda,3 Takeshi Todo,4 Gene D. Block,5,* and Wataru Nakamura1,6,*
1Laboratory of Oral Chronobiology, Graduate School of Dentistry, Osaka University, Suita, Osaka 565-0871, Japan
2Department of Life Sciences, School of Agriculture, Meiji University, Kawasaki, Kanagawa 214-8571, Japan
3Department of Mechanical Engineering, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan
4Department of Radiation Biology and Medical Genetics, Graduate School of Medicine, Osaka University, Suita, Osaka 565-0871, Japan
5Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, Los Angeles, CA 90024-1759, USA
6Precursory Research for Embryonic Science and Technology, Japan Science and Technology Agency, Kawaguchi, Saitama 332-0012,

Japan
*Correspondence: gblock@conet.ucla.edu (G.D.B.), wataru@dent.osaka-u.ac.jp (W.N.)
http://dx.doi.org/10.1016/j.celrep.2015.07.049
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

SUMMARY from gonadotropin-releasing hormone (GnRH) neurons scat-

Female reproductive function changes during aging
with the estrous cycle becoming more irregular
during the transition to menopause. We found that
intermittent shifts of the light-dark cycle disrupted
regularity of estrous cycles in middle-aged female
mice, whose estrous cycles were regular under un-
perturbed 24-hr light-dark cycles. Although female
mice deficient in Cry1 or Cry2, the core components
of the molecular circadian clock, exhibited regular
estrous cycles during youth, they showed acceler-
ated senescence characterized by irregular and
unstable estrous cycles and resultant infertility in
middle age. Notably, tuning the period length of the
environmental light-dark cycles closely to the endog-
enous one inherent in the Cry-deficient females
restored the regularity of the estrous cycles and,
consequently, improved fertility in middle age. These
results suggest that reproductive potential can be
strongly influenced by age-related changes in the
circadian system and normal reproductive func-
tioning can be rescued by the manipulation of envi-
ronmental timing signals.
INTRODUCTION
The recent trend of women having their first child at a more
advanced age (Chandra et al., 2013) has focused attention on
the physiological changes underlying irregular menstrual cycling.
In mammals, estrous cycles are not only hormonally regulated by
the hypothalamus-pituitary-gonadal (HPG) axis, but also require
an appropriately timed signal from the suprachiasmatic nucleus
(SCN) of the hypothalamus (Brown-Grant and Raisman, 1977;
Sellix and Menaker, 2010), which constitutes the central circa-
dian clock. In rodent females, the estrus cascade originates
tered throughout the basal forebrain. In the afternoon of proes-
trus, a surge of GnRH secretion into the hypophyseal portal
capillary bloodstream induces the pituitary to release a surge
of luteinizing hormone (LH) along with follicle-stimulating hor-
mone (FSH), which act on the ovary to induce ovulation and
follicular recruitment. To trigger the HPG cascade, the following
two types of input to GnRH neurons are essential: (1) estradiol
feedback from maturing ovarian follicles, and (2) a timed signal
from the SCN (Chappell, 2005; de la Iglesia and Schwartz,
2006, review). When females are entrained to a 24-hr light-dark
(LD) cycle, the phase of the LH surge is stably coupled to the
onset of locomotor activity and estrus behavior (Alleva et al.,
1971). The activity level of the locomotor rhythms of female ham-
sters vary in synchrony with the 4-day estrous cycle, regulated
by the interaction of two gonadal hormones, estradiol and pro-
gesterone (Takahashi and Menaker, 1980).
Although the causes of age-related changes in reproductive
function are complex, one factor may be a decrease in the
strength of neural signaling associated with the aging of the
SCN (Wise et al., 1988). The SCN consists of a network of
cell-intrinsic molecular circadian oscillators in which the core
components of the molecular circadian clock are Period
(Per) 1/2, Cryptochrome (Cry) 1/2, Clock, and Bmal1, which
give rise to a rhythm with a period of approximately 24 hr by
means of transcription-translation feedback loops (Mohawk
et al., 2012). Mutations of genes that are involved in these loops
can alter circadian rhythms, producing long-period, short-
period, or arrhythmic phenotypes (van der Horst et al., 1999;
Vitaterna et al., 1999). Recent studies have shown that muta-
tions of circadian clock genes can result in deficiencies of
reproductive function in young through middle-aged mice
(Miller et al., 2004; Pilorz and Steinlechner, 2008; Ratajczak
et al., 2009), indicating the importance of the daily temporal
signal for the HPG axis. The relationship, however, between
the circadian system and age-related changes in reproductive
function during the process from sexual maturation to meno-
pause remains unclear.

Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors 1407
Figure 1. Environmental Perturbation in LD Cycles Disrupts Regular Estrous Cycles in Middle-Aged WT Female Mice
(A–C) Representative double-plotted actograms of young (A) and middle-aged (B and C) female C57BL/6 mice. In the first 18 days, females were kept under a
consistent 12/12-hr LD cycle, as indicated by black and white bars above each record, then exposed to weekly shifted LD cycles for 28 days, as indicated by the
vertical line on the right.
(D–F) Daily amount of behavioral activity measured by the number of wheel revolutions corresponding to (A)–(C). Serial data for 28 days are shown before (top) and
during (bottom) the weekly shifted LD cycles, respectively. The 28-day activity data were analyzed using the periodogram to detect significance of the regularity in
estrous cycles. The p value is indicated in (D)–(F) and labeled as follows: *p < 0.05 and **p < 0.01. See also Figure S1.

RESULTS wheel-running activity were extremely regular, and pairing with

Environmental Perturbation in LD Cycles Disrupts
Regular Estrous Cycles in Middle-Aged Wild-Type
Female Mice
We divided female mice into two groups, young females aged
2–6 months and middle-aged females aged 8–12 months, and
we evaluated the regularity of their estrous cycles. Regularity
was determined by monitoring wheel-running activity (Kopp
et al., 2006). Under a 12/12-hr LD cycle, wild-type (WT) females
exhibited periodically increased nocturnal wheel-running activity
that was easily discernible on visual inspection of a double-
plotted actogram (Figure S1A). Vaginal smear cytology sampled
at zeitgeber time (ZT) 2–3 showed that spontaneous wheel-
running activity increased on the nights of proestrus, indicated
by the appearance of nucleated cells, and that estrus nocturnal
behavior, characterized by cornified cells, decreased on the
following day. The estrus was followed on the next day by metes-
trus, characterized by cornified cells, after which the appearance
of leukocytic cells for 1 or 2 days indicated the transition to
diestrus. The observed estrous cycle was consistent with that
determined by the 4- to 5-day cycle of wheel-running activity
(Figure S1A). The estrous cycles of WT females monitored by
a male mouse on nights during which activity increased resulted
in a high frequency of pregnancy.
To examine whether disharmony with the environmental cy-
cles affects regularity of estrous cycles in female mice, weekly
perturbation was applied to their environmental LD schedules
(Figure S1B). In this experiment, WT females were shifted by a
3-hr delay of darkness onset for 2 days and then returned to
the former phase of the LD cycles for 5 days, resulting in weekly
cycles. On this schedule, 100% (6/6) of young WT females (aged
4.4 months at the start of shifted cycle) displayed regular 4- to
5-day estrous cycles, which were detected by mathematical
analysis of 28-day continuous activity records of daily behavior
(number of wheel revolutions) (Figures 1A and 1D bottom). In
contrast to young females, middle-aged females showed
marked differences between activity rhythms with and without
the weekly perturbation. Among 11 of 14 middle-aged WT fe-
males (10.0 ± 1.0 months, mean ± SD) that exhibited regular
estrous cycles under normal LD cycles, 82% (9/11) showed
irregularity under the weekly shifted schedule (Figures 1B, 1C,
1E, and 1F). The difference in the ratio of regular to irregular
estrous cycles between animals on normal LD cycles and weekly
shifted cycles was statistically significant (11:3 versus 2:12,

1408 Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors
Figure 2. Middle-Aged Cry-Deficient Female Mice Exhibit Early Irregular Estrus
(A–C) Representative double-plotted actograms of WT (A), Cry1 / (B), and Cry2 / (C) female mice. Times of activity are represented by black. The same
representative young (first half, top) and middle-aged (second half, bottom) females were kept under a 12/12-hr LD cycle, as indicated by black and white bars
above each record.
(D–F) Representative estrous cycles as measured by vaginal cytology in middle-aged WT (D), Cry1 / (E), and Cry2 / (F) female mice. Two representative
individuals are shown for each genotype. N, nucleated; L, leukocytic; C, cornified.
(G) Proportions displaying significant 4- to 5-day cycles of wheel-running activity. We compared the proportions of young mice (2–6 months old) and middle-aged
mice (8–12 months old) within each genotype (Fisher’s exact test, *p < 0.05 and **p < 0.01).
(H) Regularity of estrous cycles of middle-aged females according to vaginal smear cytology. The vaginal cytology of each genotype was investigated for 3 weeks
(21 days), and the number of consecutive days on which cornified cells were evident in smears was counted. Data are represented as mean ± SE for each
genotype. See also Figure S2.

Fisher’s exact test, p = 0.002). These results indicate that envi-
ronmental perturbation of LD cycles disrupted regular estrous
cycles in middle-aged female mice.
Middle-Aged Cry-Deficient Female Mice Exhibit Early
Irregular Estrus and Resultant Infertility
Measurement of the wheel-running activity under 24-hr LD cycle
and mathematical analysis of the daily number of wheel revolu-
tions revealed that 100% (19/19) of young WT females (age
3.5 ± 0.8 months at the start of recording) displayed regular 4-
to 5-day estrous cycles (Figure 2A). Under the same condition,
79% (11/14) of young Cry1 / females (3.5 ± 0.7 months) and
79% (11/14) of young Cry2 / females (3.5 ± 1.0 months) ex-
hibited regular estrous cycles. The difference between the WT
and Cry-deficient mice was not statistically significant (Fisher’s
exact test, WT versus Cry1 / , p = 0.067, WT versus Cry2 / ,
p = 0.067). At this age, Cry-deficient females possessed the abil-
ity to reproduce (mate, become pregnant, and give birth) with no
obvious deficiency in reproductive function (Figure S2).
In contrast to young animals, during middle age (8–12 months)
marked differences emerged between WT and Cry-deficient fe-
male mice. Although 94% (16/17) of middle-aged WT females
(10.0 ± 1.0 months) exhibited regular estrous cycles, only 25%
(3/12) of middle-aged Cry1 / females (9.5 ± 1.4 months) and
8% (1/13) of Cry2 / females (9.4 ± 1.1 months) displayed regu-
larity (Figures 2A–2C). The proportions of middle-aged mice that
exhibited behavioral estrous cycles were significantly smaller
than those of young mice within the same genotype (Figure 2G).
Observations of the stages of estrus by daily vaginal smear tests
for 3 weeks (21 days) showed that middle-aged WT females
(11.3 ± 1.1 months, n = 10) exhibited regular estrous cycles, in
which cornified cells were present for 2-day consecutive periods

Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors 1409

corresponding to estrus and metestrus (Figures 2D and 2H). On
the other hand, in middle-aged Cry-deficient females, cornified
cells appeared for R3 days consecutively (Figures 2E, 2F, and
2H). These irregular vaginal smear cytologies are typically seen
during the transition to menopause at 13–16 months (Nelson
et al., 1982). When never-mated middle-age WT females
(10.2 ± 1.0 months, n = 14) were trial mated, pregnancy was
established within two trials in 71% (10/14). However, for mid-
dle-aged Cry1 / females (9.9 ± 1.5 months, n = 10) and mid-
dle-aged Cry2 / females (9.6 ± 0.4 months, n = 8), the rates
were only 10% (1/10) and 12.5% (1/8), respectively, with preg-
nancy occurring in very few mice.
Recovery from Irregular Estrus under Optimal Circadian
Periods
Next, we addressed the relationship between the circadian
rhythms and irregular estrus as well as the early infertility in
Cry-deficient female mice. We focused on the behavioral circa-
dian phenotypes of Cry-deficient mice. Similar to the previous
studies of wheel-running activity (Vitaterna et al., 1999), WT
females displayed a free-running period of 23.8 ± 0.1 hr (tWT,
1410 Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors
Figure 3. Recovery from Irregular Estrus
under Optimal Circadian Periods
(A and B) Representative double-plotted acto-
grams of middle-aged Cry1 / (A) and Cry2 / (B)
female mice. The black and white bars above and
below the recordings show the LD cycles, and
wheel-running activity rhythms for the same
representative middle-aged mice under T24h
cycle (first half, top) and either T22.5h cycle (A) or
T24.5h cycle (B) (second half, bottom) are shown.
(C and D) Representative estrous cycles as
measured by vaginal cytology in another group
of middle-aged Cry1 / (C) and Cry2 / (D) female
mice. Two representative individuals are shown
for each genotype. N, nucleated; L, leukocytic; C,
cornified.
(E) Pregnancy rates from trial matings. We
compared the pregnancy rates between young
(2–6 months) and middle-aged (8–12 months) WT
females, between middle-aged Cry1 / female
mice mated under T24h and T22.5h cycles, and
between middle-aged Cry2 / female mice mated
under T24h and T24.5h cycles (Fisher’s exact test,
*p < 0.05). See also Figure S3.
mean ± SD, n = 8), whereas Cry1 / fe-
males exhibited a short circadian period
of 22.4 ± 0.3 hr (tcry1 KO, n = 8) and
Cry2 / females a long circadian period
of 24.3 ± 0.2 hr (tcry2 KO, n = 8) (Figures
S3A–S3C). The 24-hr duration of the
12/12-hr LD-cycle environment (defined
as T24h) used as the initial experimental
condition was only slightly longer than
tWT. We considered the possibility that
the significant deviation of the period of
the LD cycle from the intrinsic periods
of the Cry-deficient mice might play a
role in altered estrous cycle regularity. We reasoned that the
atypical period difference between the environmental and the
behavioral rhythms would alter the phase relationship during
entrainment between the light cycle and the behavioral rhythm.
This abnormal phase relationship might lead to an inappropri-
ately timed circadian signal to the reproductive system.
To examine this hypothesis, we adjusted the period length of
the environmental LD cycle to a value close to the free-running
periods, tcry1 KO and tcry2 KO, for middle-aged virgin Cry-deficient
females with irregular estrous cycles (Figure 3). For middle-aged
Cry1 / females (10.7 ± 1.0 months, n = 12), the adjusted LD
cycle was 11.25/11.25 hr (T22.5h), whereas for middle-aged
Cry2 / females (10.9 ± 0.9 months, n = 13) it was 12.25/
12.25 hr (T24.5h). Wheel-running activity measurements showed
that, under these conditions, 75% (9/12) of middle-aged Cry1 /
females and 77% (10/13) of middle-aged Cry2 / females dis-
played regular estrous cycles of 4–5 days (Figures 3A and 3B).
For both Cry1 / and Cry2 / females, a significantly higher pro-
portion of middle-aged mice exhibited regular estrous cycles un-
der the adjusted LD cycle than they did under the T24h LD cycle
(Fisher’s exact test, p = 0.039 and p = 0.01, respectively). When
vaginal cytology was performed for middle-aged Cry1 /
females (10.6 ± 0.4 months, n = 6) and middle-aged Cry2 /
females (10.4 ± 0.7 months, n = 7), the regularity of estrus under
a T24h LD cycle had been improved after the mice were
switched to the corresponding T22.5h or T24.5h LD cycle (Fig-
ures 3C and 3D). Even more surprisingly, although almost no
middle-aged Cry-deficient females became pregnant under the
T24h LD cycle, pregnancy rates of 70% (7/10) and 67% (8/12)
were observed under the T22.5h and T24.5h LD cycles, respec-
tively. In both groups, increases in the pregnancy rates were sig-
nificant (Figures 3E and S3D–S3F). These results indicated that
the early appearance of irregular estrous cycles and the resultant
infertility seen in middle-aged Cry-deficient female mice could be
reversed by harmonizing the environmental and endogenous
circadian rhythms.
DISCUSSION
This study showed that optimal setting of the environmental
cycles to endogenous rhythm is one of the key factors to main-
taining regular estrous cycles in middle-aged animals. Middle-
aged Cry-deficient females, which showed early disruption of
regular estrous cycles, recovered their regularity under optimal
environmental cycles close to their intrinsic period. In middle-
aged WT mice, regular estrous cycles were observed under
the environmental cycle of a 24-hr period, implying that the
24-hr period provides their optimal circadian rhythmicity.
Weekly perturbation, which generally causes successive phase
shifts in daily activities, disturbed regularity of the estrous cycles
in the middle-aged WT females. This can be understood as the
results of disharmony between internal and environmental
rhythms. It is noteworthy that young females can maintain their
regular estrous cycles even under such inappropriate environ-
ments to a certain extent. Influence of the inharmonious lifestyle
on the estrous cycles may be much larger in middle age than in
young age.
Cry and other core clock genes (Per, Bmal1, and Clock) are ex-
pressed and give rise to the circadian rhythm in the HPG axis that
controls the estrous cycle. Since the HPG axis consists of the hy-
pothalamic GnRH neurons, the anterior lobe of the pituitary, and
the ovaries (Chappell et al., 2003; Gillespie et al., 2003; Sellix
et al., 2010), Cry-deficient mice may have impairments in these
organs. Such impairments, however, may not cause the irregular
estrous cycles evident in the middle-aged Cry-deficient females
insofar as normal reproductive function can be rescued in exper-
iments with an adjusted LD cycle. The presence of endogenous
circadian clocks at each level of the HPG axis raises the possibil-
ity that desynchronization between the circadian clock and oscil-
lations in reproductive organs rather than the misalignment of
GnRH signaling may contribute to the irregularity of reproductive
events. Reports on a circadian rhythm of ovarian sensitivity to
LH implies participation of the peripheral clock in the circadian
timing cascade of the HPG axis (Sellix et al., 2010).
The irregular estrous cycles are more likely caused by the
inability of the SCN, as the circadian pacemaker, to provide a
properly coordinated time-of-day signal to the GnRH neurons,
a similar situation to the Clock mutant females reported by Miller
et al. (2004, 2006), which involved a different core clock-gene
Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors 1411
mutation. As the master circadian pacemaker, the SCN, through
entrainment to environmental light cycles, is responsible for the
appropriate phasing of myriad physiological processes to envi-
ronmental time (Pittendrigh and Daan, 1976). The SCN-derived
signal, which is essential for the HPG axis cascade, has been
shown to be restricted to just a few hours duration in the late
afternoon of proestrus (Everett and Sawyer, 1950; Bingel and
Schwartz, 1969). Previously, phase relations between activity
onset and estrous onset were examined for hamsters exposed
to progressive T cycles. Under such T cycles, phase dissocia-
tions were evident among environmental light offset, activity
onset, and estrous onset. For longer T cycles, estrous onset
generally preceded the activity onset, while, for shorter T cycles,
it was delayed behind the running activity onset (Carmichael
et al., 1981). In Cry1 / female mice, which have the shorter
period compared with WT mice, and Cry2 / female mice with
the longer period, such phase dissociations may occur under
the T24h LD cycle. Consequently, the temporal signal may fall
outside the crucial time window for triggering the HPG cascade.
The existence of a crucial time window has been suggested
before the discovery of the SCN as a circadian pacemaker (Ever-
ett and Sawyer, 1950; Bingel and Schwartz, 1969), and, nowa-
days, the crucial time window is most probably explained as
the phase synchrony between the SCN signal and the timing of
GnRH responsiveness (Miller and Takahashi, 2013, review). In
this scheme, along with adjustments of the period of T cycles,
adjustments of the LD ratio also may be effective in achieving
the appropriate phase relationship between the crucial time win-
dow and the LD cycle. The situation is similar to the case of WT
females, which showed unstable phase angles under the condi-
tion of weekly phase shifting. The fact that young females did not
have problems with reproductive function suggests that age-
related functional changes in the SCN may be involved, although
we cannot rule out other age-related changes as well to the
reproductive axis. There is evidence that the neural output of
the SCN declines substantially with age (Nakamura et al., 2011).
The results of our study are significant in that they demonstrate
that age-related reproductive dysfunction can be reversed by
the manipulation of environmental timing signals. In humans, in
addition to the extreme environments such as jetlag and shift
work, the concept of social jetlag also has been proposed, where
the sleep/work schedule does not coincide with the sleep/wake
cycle dictated by the circadian clock (Roenneberg et al., 2012).
There may well be an association between contemporary irreg-
ular lifestyles and irregular menstrual cycles in humans. For hu-
mans, adjusting the environmental cycle period to an individual’s
intrinsic one may not be feasible as one must live with an
imposed 24-hr environmental cycle. Instead of manipulating
the environmental period length, alternative strategies such as
modifying the LD ratio may be feasible. In addition, the regulation
of light intensity may effectively shift the circadian system in such
a way that the SCN-timing signal falls within the time window to
trigger the HPG axis cascade. Favorable lighting settings, such
as increased light intensity or longer duration photoperiods,
which can restore the diminished SCN output in aging, might
be also effective for general age-related decline, such as loss
of precision and increased phase variability. Our discovery in
mice of a circadian strategy for restoring reproductive function
with altered light scheduling may provide novel approaches for
the treatment of infertility.
EXPERIMENTAL PROCEDURES
Animals
For the weekly phase-shift experiment, 10-week-old female C57BL/6J Jms
Slc mice (Japan SLC) were purchased and kept in groups until the start of
wheel-running activity measurements. Cry1+/ and Cry2+/ mice (Vitaterna
et al., 1999) were backcrossed with the C57BL/6J Jms Slc strain for ten gen-
erations and were intercrossed to generate Cry1 / and Cry2 / mice. Young
(2–6 months of age) and middle-aged (8–12 months of age) WT and homozy-
gous-deficient females were group-housed with food and water available ad
libitum. Mice were raised in the Osaka University Graduate School of Dentistry
animal facility under a 12/12-hr LD cycle (lights on at 8:00 a.m., lights off at
8:00 p.m.). Lights off was defined as ZT 12. For the trial mating, 4- to 12-
month-old male mice (C57BL/6J Jms Slc males from Japan SLC) of those
having normal reproductive ability were used. All experiments were performed
with the approval of the Committee on Animal Care and Use at the Osaka
University Graduate School of Dentistry.
Wheel-Running Activity
For experiments assessing wheel-running activity, mice were singly housed in
cages (183 3 340 3 148 mm; CL-0135, CLEA) with a running wheel (12-cm
diameter, SANKO). The cages were placed in light-tight, ventilated boxes
where the light intensity was 80 lux. The number of wheel revolutions was
counted by a magnet-sensor-activated signal between a button magnet on
the running wheel and a magnet relay (59070-010, Hamlin) fixed on a side
wall of the cage and was fed into a computer every minute. ClockLab software
(Actimetrics) was used to analyze and display the activity data.
Vaginal Cytology
A vaginal swab was collected in ZT 2–3 using a cotton-tipped toothpick wetted
with normal saline at ambient temperature. The cotton tip was inserted into the
vagina of the restrained mouse. The toothpick was gently turned and rolled
against the vaginal wall and then removed. Cells were transferred to a dry glass
slide by rolling the swab across the slide. The slide was air-dried and then
stained with 10% Giemsa stain solution (079-04391, Wako Pure Chemicals In-
dustries). The stage of the estrous cycle was assessed based on the predom-
inance of leukocytes (L), cornified epithelial (C) cells, and nucleated epithelial
(N) cells, according to Nelson et al. (1982).
Trial Mating
A single WT, Cry1 / , or Cry2 / never-mated female was paired with a single
B6 WT male overnight beginning at ZT 11–12 of the day of proestrus. Females
that did not display a clear estrous cycle were paired in a male mouse cage
for five consecutive nights, after which they were returned to their individual
cages. Females were weighed before pairing and in the mornings of day 1,
week 1, and week 2 after pairing. Pregnancy was regarded as successful if
body weight exceeded 120% of pre-pairing weight by 2 weeks after the trial
mating. Female mice with a body weight that had increased by over 120%
were sacrificed under deep anesthesia, and laparotomy was performed to
confirm the presence of fetuses. Females that had not become pregnant
were observed for at least 3 weeks after the first trial mating before a second
trial mating was performed. The pregnancy rate was calculated as the propor-
tion of mice that became pregnant by the second trial mating.
Data Analysis
A sequence of 28-day activity data was analyzed by the periodogram imple-
mented in the MATLAB Signal Processing Toolbox (MathWorks). To focus
on the estrous cycle, periodicity ranging from 3.5 to 5.5 cycles was investi-
gated. Significance of the periodicity was examined by the permutation test
(Odell et al., 1975; Ptitsyn et al., 2006), in which strength of the periodicity de-
tected in the original data was ranked among those of 105 sets of randomly
shuffled data. If the p value obtained by the ranking was less than 5%, the
periodicity was considered significant. Qualitatively similar results have been
1412 Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors
obtained in Fisher’s test (Wichert et al., 2004) as well as in chi-square periodo-
gram (Sokolove and Bushell, 1978). The methods of statistical analysis are
described in the figure legends, and data in figures are presented as the
mean ± SEM, with p < 0.05 regarded as statistically significant.
SUPPLEMENTAL INFORMATION
Supplemental Information includes three figures and can be found with this
article online at http://dx.doi.org/10.1016/j.celrep.2015.07.049.
AUTHOR CONTRIBUTIONS
W.N. designed research. N.N.T. and W.N. performed research. I.T.T. and T.T.
contributed new reagents/analytic tools. N.N.T., T.J.N., I.T.T., G.D.B., and
W.N. analyzed data. N.N.T., G.D.B., and W.N. wrote the paper.
ACKNOWLEDGMENTS
We thank Satomi Morita and Hitoshi Uchida for technical assistance. This work
was supported by the Japan Science and Technology Agency PRESTO pro-
gram (to W.N.). N.N.T. is a research fellow of the Japan Society for the Promo-
tion of Science.
Received: May 18, 2015
Revised: June 29, 2015
Accepted: July 23, 2015
Published: August 20, 2015
REFERENCES
Alleva, J.J., Waleski, M.V., and Alleva, F.R. (1971). A biological clock control-
ling the estrous cycle of the hamster. Endocrinology 88, 1368–1379.
Bingel, A.S., and Schwartz, N.B. (1969). Timing of LH release and ovulation in
the cyclic mouse. J. Reprod. Fertil. 19, 223–229.
Brown-Grant, K., and Raisman, G. (1977). Abnormalities in reproductive func-
tion associated with the destruction of the suprachiasmatic nuclei in female
rats. Proc. R. Soc. Lond. B Biol. Sci. 198, 279–296.
Carmichael, M.S., Nelson, R.J., and Zucker, I. (1981). Hamster activity and
estrous cycles: control by a single versus multiple circadian oscillator(s).
Proc. Natl. Acad. Sci. USA 78, 7830–7834.
Chandra, A., Copen, C.E., and Stephen, E.H. (2013). Infertility and impaired
fecundity in the United States, 1982-2010: data from the National Survey of
Family Growth. Natl. Health Stat. Rep. 67, 1–18, 1, 19.
Chappell, P.E. (2005). Clocks and the black box: circadian influences on
gonadotropin-releasing hormone secretion. J. Neuroendocrinol. 17, 119–130.
Chappell, P.E., White, R.S., and Mellon, P.L. (2003). Circadian gene expres-
sion regulates pulsatile gonadotropin-releasing hormone (GnRH) secretory
patterns in the hypothalamic GnRH-secreting GT1-7 cell line. J. Neurosci.
23, 11202–11213.
de la Iglesia, H.O., and Schwartz, W.J. (2006). Minireview: timely ovulation:
circadian regulation of the female hypothalamo-pituitary-gonadal axis. Endo-
crinology 147, 1148–1153.
Everett, J.W., and Sawyer, C.H. (1950). A 24-hour periodicity in the ‘‘LH-
release apparatus’’ of female rats, disclosed by barbiturate sedation. Endocri-
nology 47, 198–218.
Gillespie, J.M., Chan, B.P., Roy, D., Cai, F., and Belsham, D.D. (2003). Expres-
sion of circadian rhythm genes in gonadotropin-releasing hormone-secreting
GT1-7 neurons. Endocrinology 144, 5285–5292.
Kopp, C., Ressel, V., Wigger, E., and Tobler, I. (2006). Influence of estrus cycle
and ageing on activity patterns in two inbred mouse strains. Behav. Brain Res.
167, 165–174.
Miller, B.H., and Takahashi, J.S. (2013). Central circadian control of female
reproductive function. Front. Endocrinol. (Lausanne) 4, 195.
Miller, B.H., Olson, S.L., Turek, F.W., Levine, J.E., Horton, T.H., and Takahashi,
J.S. (2004). Circadian clock mutation disrupts estrous cyclicity and mainte-
nance of pregnancy. Curr. Biol. 14, 1367–1373.
Miller, B.H., Olson, S.L., Levine, J.E., Turek, F.W., Horton, T.H., and Takahashi,
J.S. (2006). Vasopressin regulation of the proestrous luteinizing hormone
surge in wild-type and Clock mutant mice. Biol. Reprod. 75, 778–784.
Mohawk, J.A., Green, C.B., and Takahashi, J.S. (2012). Central and peripheral
circadian clocks in mammals. Annu. Rev. Neurosci. 35, 445–462.
Nakamura, T.J., Nakamura, W., Yamazaki, S., Kudo, T., Cutler, T., Colwell,
C.S., and Block, G.D. (2011). Age-related decline in circadian output.
J. Neurosci. 31, 10201–10205.
Nelson, J.F., Felicio, L.S., Randall, P.K., Sims, C., and Finch, C.E. (1982). A lon-
gitudinal study of estrous cyclicity in aging C57BL/6J mice: I. Cycle frequency,
length and vaginal cytology. Biol. Reprod. 27, 327–339.
Odell, R.H., Jr., Smith, S.W., and Yates, F.E. (1975). A permutation test for
periodicities in short, noisy time series. Ann. Biomed. Eng. 3, 160–180.
Pilorz, V., and Steinlechner, S. (2008). Low reproductive success in Per1 and
Per2 mutant mouse females due to accelerated ageing? Reproduction 135,
559–568.
Pittendrigh, C.S., and Daan, S. (1976). A functional analysis of circadian pace-
makers in nocturnal rodents. IV. Entrainment: pacemaker as clock. J. Comp.
Physiol. 106, 291–331.
Ptitsyn, A.A., Zvonic, S., and Gimble, J.M. (2006). Permutation test for period-
icity in short time series data. BMC Bioinformatics 7 (Suppl 2), S10.
Ratajczak, C.K., Boehle, K.L., and Muglia, L.J. (2009). Impaired steroidogene-
sis and implantation failure in Bmal1-/- mice. Endocrinology 150, 1879–1885.
Cell Reports 12, 1407–1413, September 1, 2015 ª2015 The Authors 1413
Roenneberg, T., Allebrandt, K.V., Merrow, M., and Vetter, C. (2012). Social
jetlag and obesity. Curr. Biol. 22, 939–943.
Sellix, M.T., and Menaker, M. (2010). Circadian clocks in the ovary. Trends
Endocrinol. Metab. 21, 628–636.
Sellix, M.T., Yoshikawa, T., and Menaker, M. (2010). A circadian egg timer
gates ovulation. Curr. Biol. 20, R266–R267.
Sokolove, P.G., and Bushell, W.N. (1978). The chi square periodogram: its
utility for analysis of circadian rhythms. J. Theor. Biol. 72, 131–160.
Takahashi, J.S., and Menaker, M. (1980). Interaction of estradiol and proges-
terone: effects on circadian locomotor rhythm of female golden hamsters. Am.
J. Physiol. 239, R497–R504.
van der Horst, G.T., Muijtjens, M., Kobayashi, K., Takano, R., Kanno, S.,
Takao, M., de Wit, J., Verkerk, A., Eker, A.P., van Leenen, D., et al. (1999).
Mammalian Cry1 and Cry2 are essential for maintenance of circadian rhythms.
Nature 398, 627–630.
Vitaterna, M.H., Selby, C.P., Todo, T., Niwa, H., Thompson, C., Fruechte, E.M.,
Hitomi, K., Thresher, R.J., Ishikawa, T., Miyazaki, J., et al. (1999). Differential
regulation of mammalian period genes and circadian rhythmicity by crypto-
chromes 1 and 2. Proc. Natl. Acad. Sci. USA 96, 12114–12119.
Wichert, S., Fokianos, K., and Strimmer, K. (2004). Identifying periodically
expressed transcripts in microarray time series data. Bioinformatics 20, 5–20.
Wise, P.M., Cohen, I.R., Weiland, N.G., and London, E.D. (1988). Aging alters
the circadian rhythm of glucose utilization in the suprachiasmatic nucleus.
Proc. Natl. Acad. Sci. USA 85, 5305–5309.