JOURNAL
10.1177/0748730403254005
Young
Janik,
OFJanik
BIOLOGICAL
/ NONPHOTIC
RHYTHMS
PHASE
/ August
SHIFTING
2003IN FEMALE SYRIAN HAMSTERS
Nonphotic Phase Shifting in Female Syrian Hamsters:
Interactions with the Estrous Cycle
L. Young Janik and Daniel Janik1
Biology Department, University of Wisconsin–Eau Claire, Eau Claire, WI 54702
Abstract Nonphotic phase shifting of circadian rhythms was examined in
female Syrian hamsters. Animals were stimulated at zeitgeber time 4.5 by either
placing them in a novel running wheel or by transferring them to a clean home
cage. Placement in a clean home cage was more effective than novel wheel treatment in stimulating large (> 1.5 h) phase shifts. Peak phase shifts (ca. 3.5 h) and
the percentage of females showing large phase shifts were comparable to those
found in male hamsters stimulated with novel wheels. The amount of activity
induced by nonphotic stimulation and the amount of phase shifting varied
slightly with respect to the 4-day estrous cycle. Animals tended to run less and
shift less on the day of estrus. Nonphotic stimulation on proestrus often resulted
in a 1-day delay of the estrous cycle reflected in animals’ postovulatory vaginal
discharge and the expression of sexual receptivity (lordosis). This delay of the
estrous cycle was associated with large phase advances and high activity. These
results extend the generality of nonphotic phase shifting to females for the first
time and raise the possibility that resetting of circadian rhythms can induce
changes in the estrous cycle.
Key words activity, LH surge, lordosis, behavioral estrus, arousal, female hamster,
nonphotic phase shift
Nonphotic resetting of circadian rhythms is associated with behavioral activity but may be induced by
other forms of arousal such as gentle handling (Antle
and Mistlberger, 2000), refeeding (Mistlberger et al.,
1997), exposure to conspecifics (Mrosovsky, 1988), or
opportunity to hoard (Rusak et al., 1988). The role this
type of resetting plays in the biology of animals is not
clear, but evidence is accumulating that the phenomenon is widespread among mammalian species. It has
been found in various species of nocturnal rodents,
some diurnal species (ground squirrels: Hut et al.,
1999; marmosets: Glass et al., 2001), and even humans
(Buxton et al., 1997). Whatever its role, the fact that it is
found in a wide range of species, exerting large effects
in some of them, hints at the possibility that this type
of resetting is a fundamental part of the circadian system.
Nonphotic resetting has been studied extensively
using the Syrian hamster as a model organism. This
species shows peak phase advances in the middle of
the subjective day and peak phase delays late in the
subjective night (Mrosovsky et al., 1992). Furthermore, it is known that the amount of the phase shift
observed, for most circadian phases tested, is a function of the amount of the animal’s activity, usually
measured by wheel running (Janik and Mrosovsky,
1993; Bobrzynska and Mrosovsky, 1998). It has been
suggested that it is not the behavioral activity per se
that leads to nonphotic resetting, but the arousal asso-
1. To whom all correspondence should be addressed: Biology Department, UW–Eau Claire, Eau Claire, WI 54702; e-mail:
[email protected].
JOURNAL OF BIOLOGICAL RHYTHMS, Vol. 18 No. 4, August 2003
DOI: 10.1177/0748730403254005
© 2003 Sage Publications
307-317
307
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
308
JOURNAL OF BIOLOGICAL RHYTHMS / August 2003
ciated with the activity (Janik and Mrosovsky, 1993;
Biello and Mrosovsky, 1993; Antle and Mistlberger,
2000).
All of what we know about nonphotic resetting of
circadian rhythms and most of what we know about
circadian rhythms in general is based on work done in
males. However, the few studies that have examined
the circadian system of females have suggested
important sex differences. Perhaps the best example of
this comes from Zucker et al. (1980) who demonstrated that estradiol shortens the free-running circadian period of female but not male hamsters and that
this difference stems from androgen-dependent early
postnatal sexual differentiation. Because of studies
such as these, it may not be wise to extend generalizations about circadian function from males to females.
In the present article, we report that female hamsters do indeed undergo nonphotic phase shifting in a
manner that is similar to that of males. However, we
found that the specific type of stimulation that was
most effective in females may be different from that of
males and, perhaps not surprisingly, that there is some
modulation of the amount of phase shifting by the
estrous cycle. Finally, we report that the stimulation
that causes females to undergo circadian phase shifts
can result in resetting of the estrous cycle.
MATERIALS AND METHODS
General Procedure
Syrian hamsters (Mesocricetus auratus) were housed
in polypropylene bucket cages (25 × 46 × 20 cm) with
stainless steel mesh lids and running wheels 17.5 cm in
diameter. The wheels were modified by circling plastic mesh (gutter guard) around the outside, in a manner similar to that described by Mrosovsky et al.
(1998). This allowed better footing for the animals,
resulting in more wheel rotations and fewer foot injuries. A magnet glued to the running wheel triggered a
reed switch mounted to the side of the cage to monitor
wheel rotations. Switch closures were recorded using
Datacol III data collection and analysis system (Mini
Mitter, Bend, OR). Revolutions were grouped in 10min bins, and the data were not clipped. Food (Rodent
Chow 5001) and water were available at all times. The
light cycle was 14 h of light and 10 h of darkness. Light
intensity was about 300 lux measured at the level of
animals in their cages. Temperature ranged from 19 to
21 °C.
Animals’ estrous cycles were determined daily by
visually inspecting vaginal discharge (Orsini, 1961).
We followed Lisk’s (1985) description of the hamster’s
estrous cycle as indicated by vaginal discharge characteristics. On proestrus, animals have a clear, stringy
discharge. On the next day, estrus, animals have a
copious opaque and stringy discharge. This discharge
was our definitive marker to identify day in the cycle.
On the following 2 days, diestrus 1 and diestrus 2, discharge is of variable consistency and of less volume
compared to estrus.
Stimulation for nonphotic circadian clock resetting
was given at 11- to 15-day intervals. Novel wheel
exposure, one form of nonphotic stimulation we used,
was carried out in the following way. On the designated test day at ZT 4.5 (zeitgeber time; normal time of
lights off is defined as ZT 12), each female was taken
out of its home cage and placed in a novel wheel.
Novel wheels consisted of a clean wheel, of the same
type as in the home cage, mounted in a Plexiglas frame
from which the animal could not escape. Wheel revolutions were registered in home cages. Lights were
then turned off. Animals were returned to home cages
after 3 h. Animals remained in darkness for 2 more
days to assess phase shifts. During this time, the
estrous cycle was not monitored.
We used the following criteria to calculate phase
shifts. We defined activity onset as the 1st 10-min bin
of a main activity bout within 1 h of lights off with
more than 80 revolutions followed by at least 7 more
consecutive bins with at least 80 revolutions. If there
was only 1 activity bout that began within 1 h of lights
off, we used it to determine activity onset even if it was
shorter than 7 consecutive bins with 80 or more revolutions. We needed this secondary criterion only for
Experiment 1 (26% of animals and 2.6% of the tests in
Experiment 1). Phase shifts were determined by taking the difference between the average onset time for
the 2 days after the stimulus day and the average onset
time for the corresponding (same estrous days) 2 days
before the stimulus.
EXPERIMENT 1
Methods
Forty female hamsters were bred in our lab with
original stock from Harlan Spague Dawley (Madison,
WI). They had previously been tested once for activityinduced resetting at the age of 40 to 50 days old. At the
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Young Janik, Janik / NONPHOTIC PHASE SHIFTING IN FEMALE SYRIAN HAMSTERS
onset of this experiment, females were 4 to 5 months
old, except for 1 female who was 9 months old. Vaginal
discharge was monitored for 20 days to confirm regularity of cycle. Each animal was given an activity test
on each day of the cycle, but the order of testing with
respect to the particular day of the cycle was determined randomly to control for any possible order
effects. Half of the animals in each group were placed
in novel running wheels located in an adjacent room.
The other half remained in their home cages but were
also moved to the same adjacent room. After 3 h, novel
wheel–treated animals were returned to their home
cages and remained with the controls in the room for
the next 3 days.
Following the above tests, we conducted an additional experiment using only the home cage control
animals. After entrainment to LD 14:10, room lights
were turned off at ZT 4.5 and remained off for 6 days.
No other manipulations were performed.
Results
Novel wheel–treated animals showed a range of
phase shifts from 0 to over 4 h and a range of running
during the novel wheel confinement from about 500 to
7000 revolutions. However, there was no correlation
between the amount an animal ran in a novel wheel
and how much it phase shifted. Home cage controls,
on the other hand, showed a distinct pattern in the
relationship between home cage wheel running and
phase shifting. Below about 1500 revolutions, animals
showed a wide range of phase shifts from 0 to 5 h.
Above 1500 revolutions, animals showed (with 2
exceptions) large phase shifts of about 2 to 6 h. Correlations between running and shifting were statistically significant for home cage animals (proestrus r2 =
0.51, p = 0.0005; estrus r2 = 0.22, p = 0.04; diestrus r2 =
0.57, p = 0.0002; diestrus 2 r2 = 0.48, p = 0.001; Pearson
product moment correlation) but were not significant
for novel wheel–treated animals (proestrus r2 = 0.02;
estrus r2 = 0.004; diestrus 1 r2 = 0.002; diestrus 2 r2 =
0.002).
Many home cage animals and some novel wheel–
treated animals that showed large phase shifts with
fewer than 1500 revolutions in the 1st 3 h after stimulus showed a significant amount of running in the next
3 h. When total wheel revolutions registered from ZT
4.5 to ZT 10.5 were considered, the correlation coefficients with each animal’s phase shift increased: for
home cage, proestrus r2 = 0.73, p < 0.0001; estrus r2 =
0.46, p = 0.001; diestrus 1 r2 = 0.70, p < 0.0001; diestrus 2
309
r2 = 0.61, p < 0.0001. For novel wheel–treated animals,
correlations increased but not to the extent of home
cage treatment: proestrus r2 = 0.49, p = 0.0006; estrus r2
< 0.0001, p = NS; diestrus 1 r2 = 0.01, p = NS; diestrus 2 r2
= 0.46, p = 0.001. Animals tested in novel wheels on
estrus showed the poorest correlation between wheel
revolutions and phase shift.
When these data were viewed as averages, we saw
that, in general, there were no significant differences
in the amount of wheel running or phase shifting
among the 4 estrous days for either the novel wheel–
treated animals or the home cage controls (Fig. 1).
There was a tendency for wheel revolutions and phase
shifting to be lowest for animals tested on estrus, but
this was significant only for wheel revolutions measured in the novel wheel–treated animals. On average,
novel wheel–treated animals ran 2 to 3 times more in
the running wheel than home cage animals, but home
cage animals shifted almost twice as much as novel
wheel–treated animals.
Another way of looking at these data is to determine how many animals shifted above the minimal
level shown by control animals that were simply
transferred to DD at ZT 4.5. These controls showed an
average phase advance of 0.77 ± 0.70 h (mean ± SD).
Based on this, we used a conservative cutoff level of
1.5 h to distinguish between shifts that were not statistically different from controls (minimal shifts) and
those that were (large shifts). There is only a small
chance that shifts as large as 1.5 h would occur if the
home cage manipulation or novel wheel manipulation was no different in its effect than simply turning
the lights off at ZT 4.5 (p = 0.0004 in a 1-sample t test).
Using this threshold value, 9 home cage animals
showed large shifts when tested on proestrus and
estrus. Ten animals showed large shifts when tested
on diestrus 1, and 8 animals on diestrus 2. Novel
wheel–treated animals showed 4 large shifts when
tested on proestrus and estrus, 7 on diestrus 1, and 6
on diestrus 2. The proportion of animals showing
large shifts was greater in home cage animals for each
estrous day, but none of the differences was statistically significant (p > 0.05, Fisher’s exact test).
As we monitored the estrous cycle via examination
of vaginal discharge, we noticed that certain individuals showed a 1-day delay in their cycle after either the
novel wheel treatment or the home cage manipulation. The estrous cycles of 7 home cage animals and 4
novel wheel–treated animals were delayed by 1 day
immediately after the manipulation on the day of
proestrus. In fact, these animals showed the largest
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
310
JOURNAL OF BIOLOGICAL RHYTHMS / August 2003
Phase Shift, hours
(+SEM)
3
Novel Wheel
Home Cage
2
1
0
Wheel Revolutions
(x 1000,+SEM)
5
4
*
3
2
1
0
Proestrus
Estrus
Diestrus 1 Diestrus 2
Proestrus
Estrus
Diestrus 1 Diestrus 2
Test Day
Figure 1 Mean values for wheel running (ZT 4.5-7.5) and phases shifting. Asterisk indicates significant difference versus proestrus,
diestrus 1, and diestrus 2 (p = 0.001, ANOVA; p < 0.05, Bonferroni corrected t test). All other comparisons were not significant.
circadian phase shifts of all the animals tested on the
day of proestrus, all but 1 of them showing a phase
shift of at least 2.33 h. Two of these animals showed a
delay of their estrous cycle when they were manipulated on the day of diestrus 2, and they also showed
large circadian phase shifts. No animals showed an
estrous cycle delay after the novel wheel or home cage
manipulation on the days of estrus or diestrus 1.
Discussion
Female hamsters showed large circadian phase
shifts in response to novel wheel treatment with peak
shifts of about 3 to 4 h. There was not a strong tendency for females to show more phase shifting when
stimulated on any particular estrous day, but animals
did tend to run less in novel wheels when tested on the
day of estrus. Novel wheel–treated animals did not
show a correlation between the amount of running in
novel wheels and magnitude of phase shifts, but a correlation—at least for tests conducted on proestrus and
diestrus 2—did emerge when wheel revolutions dur-
ing the 1st 3 h in the home cage after novel confinement were added in.
On the other hand, control animals that were
switched to a new room in their home cage generally
showed a strong correlation between wheel running
and phase shifting whether revolutions for the 1st 3 h
or 1st 6 h after the switch were counted. Furthermore,
a greater proportion of control animals showed large
(> 1.5 h) shifts than did novel wheel–treated animals,
although the difference was not statistically significant. So what we thought would be a control procedure for novel wheel stimulation turned out to be at
least as effective in inducing phase shifts. That is the
reason we later tested animals by simply turning off
the room lights at ZT 4.5 as a further control to know
whether any phase shifts that had occurred were associated with running. That test showed an average
phase advance of 0.77 h. This amount of measured
phase shift is likely due to unmasking of activity that
occurs because animals are in LD the day before they
are given the phase-shifting stimulus or control procedure. This average control value (0.77 h) is somewhat
larger than that observed in males (e.g., approxi-
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Young Janik, Janik / NONPHOTIC PHASE SHIFTING IN FEMALE SYRIAN HAMSTERS
mately 0.5 h; Mrosovsky, 1991), most likely because
female activity rhythms are more heavily masked and
because scalloping of onsets of female activity
rhythms introduces more variability. Nevertheless, it
was clear that either novel wheel stimulation or
switching a home cage to a new room is significantly
more effective than simply placing a female into DD at
ZT 4.5.
The 1-day delay in the estrous cycle shown by some
of the females tested on proestrus was unexpected.
Prior to testing on this day of the cycle, the animals
showed a highly regular pattern of estrous cycling
almost without exception. Therefore, it is unlikely that
the delays we observed were due to any intrinsic variability in their estrous cycles.
Besides the fact that the home cage control animals
showed more phase shifting than novel wheel animals, another troubling aspect of this experiment was
the apparently low proportion of females that showed
large phase shifts. It seemed possible that the number
of shifters was somewhat low because the animals
were about 7 months old by the end of the experiment—quite a bit older than is typically used in circadian experiments. Because of these doubts, we
decided to repeat the tests with some differences. First,
we used younger animals—about 65 days old at the
beginning of testing. Second, we used a within-subjects
design to make a direct comparison between individuals as to whether novel wheel treatment or home
cage treatment was more effective. Third, we modified
our treatment of home caged animals: they were first
moved to the experimental room and then their cages
were changed.
EXPERIMENT 2
Methods
Twenty-four female hamsters were obtained from
Harlan Sprague Dawley at 33 days of age. Prior to testing, they were housed in LD for 31 days. We used a different type of mesh (plastic-coated polyester) on running wheels in cages and in novel wheels than in
Experiment 1. We found that this type of mesh was
finer and had no rough edges, and its use resulted in
fewer abrasions.
In this group, all animals were given activity tests
with novel wheels on each of their estrous days. They
were tested all at once, and the order of the tests with
respect to the animals’ estrous day was random. Sub-
311
sequently, these animals were given another series of 4
tests, each on a different estrous day, but they were
given fresh running-wheel cages (instead of novel
wheels) and lights were turned off at ZT 4.5. For both
series of tests, animals were moved into an adjacent
room.
Results
Home cage animals and novel wheel animals generally showed a significant correlation between wheel
revolutions produced during the test and the resulting
phase shifts (Fig. 2). Correlations were generally
higher for home cage treatment on any given estrous
day except for the day of proestrus on which both
treatments resulted in a comparably high correlation.
When tests on all estrous days were considered, 16
of 23 animals showed a large (> 1.5 h) phase shift in
response to the home cage treatment, whereas 12 of 24
showed a large phase shift in response to novel wheel
treatment. Comparison of home cage treatment with
novel wheel treatment for particular estrous days generally showed that more animals shifted in response to
home cage treatment. The difference was statistically
significant on diestrus 1 (p = 0.017, Fisher’s exact test).
When animals were tested with novel wheels on
proestrus, 4 of them showed a 1-day delay of their
estrous cycle (Fig. 2). When tested with home cages on
proestrus, 12 animals showed a 1-day delay. All but 2
of the phase shifts for these animals were greater than
1.5 h. Estrous cycling was not altered for any animal
after testing on any other day of the cycle.
Female hamsters entrained to LD 14:10 showed a
robust and predictable 4-day pattern of wheel running
corresponding with their estrous cycle (Fig. 3A).
When we first introduced the animals to wheel cages,
activity levels were uniformly high across all estrous
days. Several weeks later as overall activity levels
tapered off, a clear pattern emerged. Nightly activity
levels were highest on the day of proestrus and
diestrus 2, lowest on estrus, and intermediate on
diestrus 1. Animals that had a 1-day delay of their
estrous cycle after nonphotic testing displayed what
appeared to be 2 consecutive days of proestrus-like
activity—on the day of the nonphotic stimulation and
on the day after nonphotic stimulation (Fig. 3B).
Discussion
The results of this experiment with younger animals generally confirmed the outcomes of the 1st
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
312
JOURNAL OF BIOLOGICAL RHYTHMS / August 2003
Figure 2. Activity response relations for each estrous day in Experiment 2. Wheel revolutions were counted from ZT 4.5 to ZT 7.5 (time in
novel wheels or equivalent time for home cages). Open circles: phase shifts of females with estrous cycle on schedule. Solid circles: phase
shifts of females with delayed postovulatory discharge after testing. Dashed lines indicate the distinction between large shifters and minimal shifters. p values are for Pearson product moment correlation.
series of tests. Both series indicated that home cage
treatment resulted in more animals showing large
phase shifts. Home cage treatment also resulted in a
better correlation between wheel revolutions and the
subsequent phase shift. In Experiment 1, 63% of the
home cage animals showed large phase shifts in one or
another of the tests, whereas in Experiment 2, 70% did.
However, home cage manipulation was conducted
differently for each series. In the 1st series, animals
were switched to an adjacent room but remained in
the same home cages. In the 2nd series, in addition to
being moved to the experimental room, they were
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Young Janik, Janik / NONPHOTIC PHASE SHIFTING IN FEMALE SYRIAN HAMSTERS
313
formed in Experiments 1 and 2 to determine which
was more effective in producing large nonphotic
shifts. Second, we aimed to determine how the 1-day
delay in the estrous cycle occurred. Did animals show
2 consecutive days of heat as suggested by the observation that daily activity (alpha) appeared proestruslike on the day of the test and on the day after the test?
Or did they skip the 1st period of heat that would have
begun several hours after the cage change and then go
into heat the day after?
EXPERIMENT 3
Methods
Figure 3. Estrous modulation of circadian activity. (A) Mean
wheel revolutions over 24 h for the 4 days of the estrous cycle.
Data were collected over a 4-day period when the hamsters were
15 to 16 weeks old. (B) Activity recording of a female hamster
given a cage change on proestrus. Light cycle prior to cage change
is diagrammed at bottom. This female had a phase advance of
about 3 h. Note that activity level on the day after the cage change
is like that on proestrus (high), and activity 2 days after the cage
change is like that on estrus (low).
switched to fresh home cages. The fact that the stimulus was given somewhat differently, yet the outcomes
were similar, attests to the robustness of the
phenomenon.
As in the 1st experiment, estrous cycling was
delayed in animals that showed large circadian phase
shifts on the day of proestrus. More animals showed
this delay than in Experiment 1 corresponding with
the greater number of large shifts in this experiment.
Unlike Experiment 1, none of the animals showed a
delay in their estrous cycle in response to stimulation
on any day but proestrus.
We conducted a 3rd series of tests to resolve 2
remaining questions. First, we made a direct comparison between the different home cage treatments per-
Twenty-eight female hamsters, 39 days old, were
obtained from Harlan Sprague Dawley. Housing, care,
and estrous cycle assessment were the same as in the
previous experiments. Animals were held in LD for 23
days prior to the 1st test. On the day of the test at ZT
4.5, half of the hamsters remained in their home room
and half were transferred to an adjacent room. All of
the animals were then given fresh cages, and lights
were turned off and remained off for the next 3 days.
Animals were tested only on estrus, diestrus 1, or
diestrus 2. They were not tested on proestrus because
an estrous delay would have led to logistical problems
in carrying out the next part of the experiment. After
re-entrainment 14 days later, the procedure was
repeated with animals assigned to the opposite treatment group. Individuals were given both treatments
on the same day of their estrous cycle.
To test for heat, the same females were used. After
re-entrainment for about 30 days, females were
moved into the experimental room at ZT 4.5 on the
day of proestrus. They were given fresh cages, and
lights were turned off. At ZT 12.5, each female was
removed from her home cage and placed in an empty
bucket cage with a male hamster. Using a safe light,
the pair was observed for 3 min. If a female showed a
fixed posture and sustained it for the remainder of the
3-min period, we considered it lordosis (during the
testing, we found that lordosis was immediate).
Mounting was not permitted. The following day at ZT
12.5, each female was again tested with a male in the
same way. After 3 days, the light-dark cycle was
reestablished.
As in Experiment 1, a final control procedure was
conducted. Without moving or cage changing, the
lights were turned off at ZT 4.5 and remained off for
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
314
JOURNAL OF BIOLOGICAL RHYTHMS / August 2003
change), they showed a mean phase advance of 0.48 ±
0.60 h (mean ± SD). This value is statistically different
from 0.80 h in 1 sample t test (p = 0.03).
Discussion
Figure 4. Experiment 3. Phase shifts of female hamsters as a function of running in fresh home cages on proestrus. Open circles:
phase shifts for females showing lordosis on the evening of
proestrus. Closed triangles: phase shifts of females that did not
show lordosis on the day of the cage change but did show it the
next day.
the next 10 days and animals were assessed for phase
shifting as described previously.
Results
Both treatments—moving the home cage from one
room to an adjacent room then changing the cage, and
giving animals a fresh home cage in the same room—
resulted in similar amounts of phase shifting. When
animals were moved to an adjacent room 15 of 28
phase shifts were ≥ 1.5 h, and when they stayed in the
same room, 13 out of 28 phase shifts were ≥ 1.5 h (p >
0.05, Fisher’s exact test).
Figure 4 shows the activity response relation on
proestrus with the subsequent lordosis test. Seven of
the 10 animals with large phase shifts did not show
lordosis on the day of the test but did the day after. All
animals with minimal shifts showed lordosis the day
of the test and no receptivity the day after. Postovulatory discharge was as predicted from the behavioral
tests: the 7 females with delayed lordosis showed
delayed postovulatory discharge. There was an exception: 1 female showed lordosis the day of the proestrus
test, but her postovulatory discharge was delayed 2
days. She showed an intermediate number of wheel
revolutions in the 1st 3 h after the cage change (3678)
and phase shifted 2.17 h.
When this group of animals was put into DD at ZT
4.5 while in their home cage (no fresh cage, no room
The results of the 3rd group of tests indicated it is
not particularly important whether females are transferred from one room to another in addition to changing their cage, or from one cage to another in the same
room to achieve large nonphotic phase shifts. Both
procedures resulted in a percentage of animals showing large shifts and in maximal shifts that were comparable to each other and to the previous experiments.
This outcome is consistent with the findings of Galani
et al. (2001) who showed that there was no additive
effect of 2 stimuli—cage changing or movement to an
unfamiliar room—on the amount of locomotor activity of rats.
The control manipulation performed in this series
of tests, in which animals were transferred into DD at
ZT 4.5 without a cage change or room change, confirmed that most observed phase shifts of 1.5 h or more
using our procedures are due to something other than
the light-to-dark transition. Indeed, 1.5 h is a conservative threshold to use to differentiate between the
“noise” generated by the phase-shifting procedure
and the specific effects of a fresh cage or a novel wheel.
The true limit of the noise level is most likely in the
neighborhood of 0.8 to 1.2 h. These values were the
upper 95% confidence intervals of the 2 groups of
females we tested (Experiments 1 and 3).
This raises the general issue of why we chose to test
females for nonphotic phase shifting using a procedure in which they are stimulated at or soon after the
transition from light to dark—a modification of
Aschoff’s (1965) type 2 procedure—as opposed to a
more traditional DD free-run procedure. This procedure has several advantages that have been addressed
extensively elsewhere (Mrosovsky 1996). Among
them, it allows the experimenter to obtain an accurate
estimate of phase shift in only a couple of cycles
because there are no transients. Measurements
obtained in this way correspond closely with those
obtained by the traditional DD free-run method.
Additionally, there were several constraints intrinsic
to the system we were studying that made it difficult
to assess phase shifting in DD. First, activity (and perhaps the arousal associated with it) dwindles as animals remain in DD for prolonged periods. Since we
were stimulating animals up to 4 times, this would
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Young Janik, Janik / NONPHOTIC PHASE SHIFTING IN FEMALE SYRIAN HAMSTERS
have meant keeping animals in DD for longer than 2
months, or if we had brought animals into LD for reentrainment between stimulations, the period of time
between stimulations would have been greatly
increased. Second, not all animals show estrous
cyclicity reflected in their activity records, so it was
essential to get verification of their estrous day via
visual inspection of vaginal secretions. Third, in DD,
animals will drift out of (circadian) phase with each
other. Conducting individual tests at different clock
times for each animal would be disruptive to the animals not being tested at the time. Fourth, we wanted to
compare our data with data from nonphotic phaseshifting experiments in males, many of which have
used this procedure. Probably the main disadvantage
of using this procedure is that female hamsters show a
fair amount of masking so that there is a fairly large
apparent phase shift when animals are transferred
from LD to DD.
GENERAL DISCUSSION
Female hamsters show robust nonphotic resetting
of circadian rhythms similar to that found in males.
Although we do not know what the phase response
curve for nonphotic stimulation looks like in females,
our results for stimulation beginning at ZT 4.5 correspond with previous results obtained using male
hamsters (Mrosovsky et al., 1992; Bobrzynska and
Mrosovsky, 1998). At this phase, maximal shifts in
both sexes are about 3.5 h and the activity-response
characteristics are similar with both sexes showing
large phase advances when they run in a novel wheel
around 4000-6000 revolutions or greater. The percentage of males that show phase advances of about 1.5 h
or greater (57%: Bobrzynska and Mrosovsky, 1998;
47%: Janik and Mrosovsky, 1993) is comparable to the
percentages we found in the present study, which varied from 20% to 53% in the 1st experiment, 17% to 52%
in the 2nd experiment (Fig. 2), and 54% in the 3rd
experiment.
The percentage of females showing large phase
shifts varied depending on the type of nonphotic stimulus given. Females showed greater phase shifting in
response to either a fresh home cage or having their
home cage moved to a new room as compared to
placement in a novel wheel (Figs. 1, 2). Work with
male hamsters has shown the opposite; they produce
greater phase shifts in response to novel wheels than
to changing their home cage (for example, compare
315
Mrosovsky [1988] to Reebs and Mrosovsky [1989]).
The situation in females is interesting because they
showed more wheel running in novel wheels than in
home cages yet they shifted more after running in the
home cages than in the novel wheels. This suggests
that being in a fresh cage or a new room is more arousing to a female than is a novel wheel and that behaviors associated with cage changing such as exploration, scent marking, food hoarding, and nest building,
as well as wheel running, are manifestations of
arousal in female hamsters. Our (unquantified) observations of hamsters after changing their cages would
indicate that this is a possibility, as the females showed
many of these behaviors. Despite the possibility that
activities other than running in an exercise wheel may
be a stimulant for phase shifting, wheel running in a
home cage after a change is a good predictor of subsequent phase shifting. In fact, it is a better predictor of
phase shifting than novel wheel running, as indicated
by the strong correlations we saw (Fig. 2).
The amount of phase shifting shown by female
hamsters also varied as a function of the estrous day
on which they received nonphotic stimulation. The
effect was modest, but stimulation on the day of estrus
consistently produced the lowest number of large
phase shifts and the least running. This variation in
induced running and phase shifting corresponds with
the amount of clock-controlled nighttime running
female hamsters show over the estrous cycle. The
present study (Fig. 3) and several other studies using
other means of measuring activity (Richards, 1966;
Morin et al., 1977; Takahashi and Menaker, 1980) have
shown that diestrus 2 and proestrus are the days of
highest activity and that female hamsters show the
least activity on estrus.
The most striking and unanticipated result in the
current study was the delay in the estrous cycle we
observed after nonphotic stimulation on the day of
proestrus. There is strong evidence that the estrous
cycle of hamsters is tightly controlled by and coupled
to the circadian system (Alleva et al., 1971; Fitzgerald
and Zucker, 1976; Carmichael et al., 1981; Swann and
Turek, 1985). Therefore, the present data suggesting
that stimulation that leads to an advance of the circadian system can also lead to a delay in the estrous cycle
is remarkable.
Three pieces of evidence demonstrate that the
estrous cycle of some females is immediately delayed
after nonphotic stimulation on the day of proestrus.
First, postovulatory discharge was delayed by 1 day
from the predicted day after the animals were brought
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
316
JOURNAL OF BIOLOGICAL RHYTHMS / August 2003
into LD. Second, activity records of these animals
showed a 2nd consecutive night of high (proestruslike) activity after the pulse day followed by a night of
low (estrus-like) activity. Third, in Experiment 3, we
found that animals that were subsequently found to
have a delayed postovulatory discharge did not display lordosis later on the day of the nonphotic stimulation, but they showed it about 24 h later. It is unlikely
that the estrous delays we observed were spontaneous
or random. During all of our testing, only a couple animals displayed irregularity in their estrous cycles.
Regular cage changing and light cycle manipulations
for testing did not affect estrous cycling.
Lordosis behavior is dependent on the proestrus
LH surge (Bosley and Leavitt, 1972; Goldman and
Sheridan, 1974). Therefore, our data suggest that
females not showing lordosis on the day of proestrus
did not have the predicted LH surge. The proestrus
LH surge in hamsters takes place over a 6-h period
beginning about 4 h before lights off, corresponding to
ZT 8 to ZT 14. The peak is about 2 h before lights off.
Alleva and Umberger (1966) and Siegel et al. (1976)
have shown that certain drugs, such as phenylisopropylhydrazine and phenobarbital, block the LH
surge and ovulation if given before about ZT 8 on the
day of proestrus. The surge and ovulation then takes
place about 24 h later (Lippman, 1968; Stetson and
Watson-Whitmyre, 1976; Siegel et al., 1976; Morin,
1979). The LH surge can be delayed repeatedly in this
way for up to 3 days in a row, the delayed surge always
appearing at the same circadian phase (Stetson and
Watson-Whitmyre, 1977). This, and the observation
that the daily LH surge in estrogen-treated
ovariectomized hamsters splits in parallel with locomotor activity after females are exposed to prolonged
constant light, demonstrates that the LH surge is
linked to circadian timing (Swann and Turek, 1985).
Is there a connection between the action of drugs
that delay the LH surge and the action of nonphotic
stimulation? Certainly the circadian phases at which
the drugs were administered (ZT 6) and at which animals were nonphotically stimulated in the present
study (ZT 4.5) correspond closely. Alleva and
Umberger (1966) noted that drugs that delay ovulation also induce “behavioral excitement” (they did not
define what they meant by this), whereas drugs that
were ineffective in blocking ovulation depressed
behavioral activity. If it is indeed true that drugs effective in delaying the LH surge on proestrus also stimulate behavioral activity, then it is possible that they are
engaging the same mechanisms as nonphotic stimula-
tion. This raises the issue of whether it is behavioral
activation (arousal) associated with nonphotic stimulation or the circadian resetting that results from the
stimulation that causes the delay of the estrous cycle.
Certainly, the majority of animals that showed the
estrous delay showed large circadian phase shifts, but
they also tended to be the animals that showed the
most activity.
These facts suggest a mechanism that explains how
the behaviorally induced delay in the present study
occurs. If at ZT 4.5, a female runs sufficiently to cause a
rapid 3-h circadian advance (and there is evidence that
suggests that nonphotic advances are rapid;
Maywood et al., 1999), that would bring its circadian
time to about ZT 10.5 at the end of the 3-h pulse. That
is, the phase advance would essentially cause the animal’s circadian timing system to skip the phase at
which the first part of the LH surge occurs, beginning
at about ZT 8.
If this hypothesis were correct, it would be interesting to note whether drugs that cause a delay in the LH
surge also induce behavioral activity or cause circadian phase advances. Also, the fact that drugs that
block the proestrus LH surge when administered
before ZT 7 do not block the LH surge when given at
later ZTs suggests that cage changes that begin about
ZT 8 (assuming they cause significant circadian
advances) will not cause estrous delays. If so, there
may be a common underlying mechanism.
ACKNOWLEDGMENTS
This work was supported by grant IBN98066 from
the National Science Foundation. We wish to thank
Gwendolyn Applebaugh and William Applebaugh
for their suggestions on experimental design and 3
anonymous reviewers for suggestions to improve the
manuscript.
REFERENCES
Alleva JJ and Umberger EJ (1966) Evidence for neural control of the release of pituitary ovulating hormone in the
Golden Syrian hamster. Endocrinology 78:1125-1129.
Alleva JJ, Waleski MV, and Alleva FR (1971) A biological
clock controlling the estrous cycle of the hamster. Endocrinology 88:1368-1379.
Antle MC and Mistlberger RE (2000) Circadian clock resetting by sleep deprivation without exercise in the Syrian
hamster. J Neurosci 20:9326-9332.
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Young Janik, Janik / NONPHOTIC PHASE SHIFTING IN FEMALE SYRIAN HAMSTERS
Aschoff J (1965) Response curves in circadian periodicity. In
Circadian Clocks, J Aschoff, ed, pp 95-111, North-Holland,
Amsterdam.
Biello SM and Mrosovsky N (1993) Circadian phase shifts
induced by chlordiazepoxide without increased locomotor activity. Brain Res 622:58-62.
Bobrzynska K and Mrosovsky N (1998) Phase shifting by
novelty-induced running: activity dose response curves
at different circadian times. J Comp Physiol A 182:251258.
Bosley LG and Leavitt WW (1972) Dependence of
preovulatory progesterone on critical period in the cyclic
hamster. Am J Physiol 222:129-133.
Buxton OM, Frank SA, Littermite-Baleriaux M, Leproult R,
Turek FW, and Van Cauter E (1997) Roles of intensity and
duration of nocturnal exercise in causing phase delays of
human circadian rhythms. Am J Physiol 273:E536-E542.
Carmichael M, Nelson R, and Zucker I (1981) Hamster activity and estrous cycles: control by a single versus multiple
circadian oscillator(s). Proc Natl Acad Sci U S A 78:78307834.
Fitzgerald K and Zucker I (1976) Circadian organization of
the estrous cycle of the golden hamster. Proc Natl Acad
Sci U S A 73:2923-2927.
Galani R, Duconseille E, Bildstein O, and Cassel JC (2001)
Effects of room and cage familiarity on locomotor activity
measures in rats. Physiol Behav 74:1-4.
Glass JD, Tardif SD, Clements R, and Mrosovsky N (2001)
Photic and nonphotic circadian phase resetting in a diurnal primate, the common marmoset. Am J Physiol
280:R191-R-197.
Goldman BD and Sheridan PJ (1974) The ovulatory surge of
gonadotropins and sexual receptivity in the female
golden hamster. Physiol Behav 12:991-995.
Hut RA, Mrosovsky N, and Daan S (1999) Nonphotic
entrainment in a diurnal mammal, the European ground
squirrel (Spermophilus citellus). J Biol Rhythms 14:409-419.
Janik D and Mrosovsky N (1993) Nonphotically induced
phase shifts of circadian rhythms in the golden hamster:
activity-response curves at different ambient temperatures. Physiol Behav 53:431-436.
Lippman W (1968) Relationship between hypothalamic
norepinephrine and serotonin and gonadotrophin secretion in the hamster. Nature 218:173-174.
Lisk RD (1985) The estrous cycle. In The Hamster: Reproduction and Behavior, HI Siegel ed, pp 23-51, Plenum Press,
New York.
Maywood ES, Mrosovsky N, Field MD, and Hastings MH
(1999) Rapid down-regulation of mammalian period
genes during behavioral resetting of the circadian clock.
Proc Natl Acad Sci U S A 96:15211-15216.
Mistlberger RE, Sinclair SV, Marchant EG, and Neil L (1997)
Phase shifts to refeeding in the Syrian hamster mediated
by running activity. Physiol Behav 61:273-278.
317
Morin LP (1979) Effect of ovarian hormones on synchrony of
hamster circadian rhythms. Physiol Behav 24:741-749.
Morin LP, Fitzgerald KM, and Zucker I (1977) Estradiol
shortens the period of hamster circadian rhythms. Science 196:305-307.
Mrosovsky N (1988) Phase response curves for social
entrainment. J Comp Physiol A 162:35-46.
Mrosovsky N (1991) Double-pulse experiments with
nonphotic and photic phase shifting stimuli. J Biol
Rhythms 6:167-179.
Mrosovsky N (1996) Locomotor activity and non-photic
influences on circadian clocks. Biol Rev 71:343-372.
Mrosovsky N, Salmon PA, Menaker M, and Ralph MR (1992)
Nonphotic phase shifting in hamster clock mutant. J Biol
Rhythms 7:41-49.
Mrosovsky N, Salmon PA, and Vrang N (1998) Revolutionary science: an improved running wheel for hamsters.
Chronobiol Int 15:147-158.
Orsini MW (1961) The external vaginal phenomena characterizing the stages of the estrous cycle, pregnancy,
pseudopregnancy, lactation, and the anestrous hamster.
Proc Anim Care Panel 11:193-206.
Reebs SG and Mrosovsky N (1989) Effects of induced wheelrunning on the circadian activity rhythms of Syrian hamsters: entrainment and phase response curve. J Biol
Rhythms 4:39-48.
Richards MPM (1966) Activity measured by running wheels
and observation during the oestrous cycle, pregnancy
and pseudopregnancy in the golden hamster. Anim
Behav 14:450-458.
Rusak B, Mistlberger RE, Losier B, and Jones CH (1988)
Daily hoarding opportunity entrains the pacemaker for
hamster activity rhythms. J Comp Physiol A164:165-171.
Siegel HI, Bast JD, and Greenwald GS (1976) The effects of
phenobarbital and gonadal steroids on periovulatory serum levels of luteinizing hormone and follicle-stimulating
hormone in the hamster. Endocrinology 98:48-55.
Stetson MH and Watson-Whitmyre M (1976) Nucleus
suprachiasmaticus: the biological clock in the hamster?
Science 191:197-199.
Stetson MH and Watson-Whitmyre M (1977) The neural
clock regulating estrous cyclicity in hamsters: gonadotropin release following barbiturate blockade. Biol Reprod
16:536-542.
Swann JM and Turek FW (1985) Multiple circadian oscillators regulate the timing of behavioral and endocrine
rhythms in female golden hamsters. Science 228:898-900.
Takahashi JS and Menaker M (1980) Interaction of estradiol
and progesterone: effects on circadian locomotor rhythm
of female golden hamsters. Am J Physiol 239:R497-R504.
Zucker I, Fitzgerald KM, and Morin LP (1980) Sex differentiation of the circadian system in the golden hamster. Am J
Physiol 238:R97-R101.
Downloaded from jbr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016