Preferences of mice, Mus musculus, for different
types of running wheel
s.
Banjanin & N. Mrosovsky
Department of Zoology, University of Toronto, 25 Harbord Street, Toronto, Ontario, Canada M5S 3G5
Summary
Mice are increasingly used in research. In particular, their wheel running is often used as a
measure of activity, and as a marker of phase of circadian rhythms. Learning about the
preferences of mice for different types of wheel may improve their welfare and suggest ways of
increasing activity levels. Mice, Mus musculus, were given a choice between different types
of running wheel by putting them in cages equipped with two wheels. Strong preferences
were shown for wheels with a plastic mesh flooring, rather than the standard metal rods only.
The mesh was even preferred over a solid base, although this effect was not seen in mice that
had been given access only to wheels with the solid base immediately prior to the choice test.
Small diameter wheels, sometimes sold as mouse wheels, were preferred less than standardsized wheels with rods. The results suggest that types of running wheel often used in
laboratories can be improved by considering the animals' preferences. The types of wheel
tested here are easy to maintain and entail little additional cost, while increasing the mouse/s
interest in running and exercise.
Keywords
Mice; preferences; rhythms; wheel running
Running in a wheel is generally rewarding;
many species engage voluntarily in this
activity (Sherwin 1998a). More time is spent
in wheel running than in alternative activities, such as moving around in adjacent
areas or exploring a complex maze (Brant &
Kavanau 1964/ Sherwin 1998b). Further
indications that wheel running is a rewarding
experience are bar pressing to obtain access
to a wheel (Collier et al. 19901, returning to a
place where previously there had been an
opportunity to run in a wheel (Antoniadis &
McDonald 1997)/ and effectiveness of wheels
as a reinforcer for visual discriminations
(Iversen 1998).
Why animals run in wheels, and how this
relates to behaviour occurring in natural circumstances/ are complicated questions. The
present experiments are not concerned with
Correspondence
Accepted
to: N. Mrosovsky
11 January
2000
©
these matters, but have the more modest aim
of learning more about what type of running
wheel might be appropriate for research use
and welfare. Mice were studied because their
wheel running is widely used as a marker of
cycle phase in experiments on biological
rhythms (Moore-Ede et al. 1982)/ and more
generally because the availability of transgenic and mutant strains makes this species
increasingly valuable in many types of
research.
Two pieces of information from investigations with hamsters guided the present studies. First, standard rod wheels (diameter
17.5 cm) are preferred to small wheels (diameter 13.0 em). Second, placing a plastic
mesh around the floor of the running wheel
greatly increases the number of revolutions
made (Mrosovsky et al. 1998). In the present
experiments, we investigated the effect of
these variables, wheel size and type of floor,
in mice.
Laboratory
Animals
Ltd. Laboratory
Animals
(2000) 34, 313-318
Banjanin & Mrosovsky
314
Materials and methods
Animals
and environment
Male C57BL/6 mice were housed in polypropylene cages (45.5 x 37.5 x 22.5 em),
equipped with two running wheels (a single
animal per cage). The choice of wheels was
varied in different experiments, but in all
cases the base of the wheels was 2.5 em off
the floor. Heat-treated, hardwood laboratory
bedding was provided (Beta Chip, Northeastern Products Corp.). Food pellets (Lab
Diet, 5001 Rodent Diet, PM! Nutrition
International, Inc.) and tap water were
available ad libitum, and were placed in the
cage equidistant from each wheel. The animals were kept in a 12:12 h light-dark (LD)
cycle, with lights off at 19:00 hi artificial
illumination during the light portion of the
LD cycle, as measured with an ISO-TECH
ILM350 meter, was about 100 lux in Experiments I, 4 and 5 and 100-600 lux in Experiments 2 and 3. Room temperature was 20
± 4°C. Revolutions were recorded using
Dataquest hardware and software.
Experimental
procedure
The animals were put in the wheel-choice
cages at 14:00 h on Day 1. They were given a
day to adjust to the surroundings and data
collection began at 12:00 h on Day 2. On Day
7, the two wheels were interchanged, to
control for possible position preferences. The
experiments ended on Day 13. Thus, wheel
revolutions were recorded for 5 days before
and 5 days after the position of the wheels
was switched.
Before and after each experiment, the legs
of the mice were inspected. No calluses or
sore areas were detected in any of the
experiments, and this is not considered further here.
Experiment
mesh
diameter of about 17.5 em, a width of 7.5 em,
and a floor composed of 1.6 mm diameter
metal rods spaced 12.0 mm apart. The other
wheel was the same, except that in addition
it had a black plastic mesh (approx.
3.5 x 3.5 mm mesh) wrapped around the
outside of the wheel (Fig 1). The animals still
made contact with the rods, but the mesh
outside provided an additional surface. Once
attached, the plastic mesh remained in place,
and could be put through a cage washing
machine_
Experiment
2: Plastic mesh vs solid floor
Twelve mice, 90 days old, originally obtained
from Harlan-Sprague Dawley (Indiana), were
studied. Of these, seven had no previous
experience with wheels, and five had been
with standard wheels with rods (see above)
for about 3 weeks, ending 20 days before the
present experiment. The mice were given a
choice between wheels covered with plastic
mesh, as used in Experiment I, and wheels
with a solid transparent flexible strip of
plastic covering the outside of the wheel, in
the same manner as the plastic mesh (Fig 1).
Experiment 3: Standard rods vs small
wheels with rods
Eight mice, 90-120 days old, originally
obtained from Harlan-Sprague Dawley (Indiana), were studied. Of these, four had previously been in a standard wheel with rods
for 3-8 weeks. In the two-wheel preference
test , one wheel was the standard wheel type
1: standard rods vs plastic
Six mice, 340-400 days old, were obtained
from the breeding colony in the Department
of Zoology, University of Toronto. None had
experience with running wheels before. In
the present experiment, one of the wheels
was the standard commercially available
wheel (Hagen, Montreal, Quebec), with a
Laboratory Animals (2000) 34
Fig 1 From left to right, standard wheel with rods,
wheel with mesh cover, and wheel with solid plastic
cover
315
Running wheels for mice
with rods used in Experiment I, i.e. 17.5 cm
in diameter; the other wheel was only
13.0 cm in diameter and 6.0 cm wide, but the
rods were the same size and distance apart as
those in the standard wheel.
Experiment 4: Effect of experience in
wheels with solid floors
Eight mice, 210 days old, originally obtained
from Harlan-Sprague Dawley (IndianaL were
studied. Two out of these eight had 3-8
weeks previous experience with a standard
wheel with rods. The mice were first placed
in polypropylene cages (44.0 x 23.0 x 20.0 cm)
equipped with a single running wheel. This
wheel was of the standard, 17.5 cm diameter,
type and surrounded by the solid flooring as
described for Experiment 2. After 7-14 days
they were transferred to cages holding two
wheels, and given a choice between solidcovered and mesh-covered wheels.
Experiment 5: Activity levels with single
wheels
This experiment differed from the previous
ones in that the mice were kept in smaller
cages (44.0x 23.0 x 20.0 cm) and only one
wheel (17.5 cm diameter) was present. Ten
animals had standard wheels with rods, and
11 had wheels with plastic mesh, as in
Experiment 1. Six days after arrival (at 56
days old) from Charles River, Quebec, they
were placed in the cages with running
wheels, and kept there for 17 daysi the first
1.5 days were excluded from the data analysis
because of problems with recording.
mean and SEM, P < 0.0001, Figs 2 and 3). The
mesh base was also preferred to a solid floor
(Experiment 2, 71.6±6.6%, mean and SEM,
P=0.007, Figs 2 and 3); there was no significant difference in the preferences of mice
with previous experience in wheels and those
encountering wheels for the first time in this
experiment (P = 0.2, two-tailed t-testl.
The standard rod wheel was preferred to
the small one (Experiment 3, 69.7± 7.8%,
mean and SEM, P = 0.04, Figs 2 and 4li again
there were no differences between wheelnaive mice and those with previous experience in wheels (P = 0.8, two-tailed t-test). It
should be noted that if one considers that the
circumference of the smaller wheel was only
0.74 times that of the larger wheel, in terms
of distance covered, the preference (75.7%)is
stronger than the 69.7% calculated in terms
of revolutions.
One to 2 weeks of experience with a solid
wheel floor immediately prior to a choice test
between such a wheel and one covered with
mesh appeared to abolish the preference seen
in the other experiments for the mesh design
(Experiment 4, 55.0±8.8%, mean and SEM,
for the mesh, P=0.59, Figs 2 and 4).
The results of Experiment 5, with only one
wheel available, agreed with those of the
choice tests, in that the number of revolutions was consistently much higher when
..-- 100
~~'"
0"0
0
-.c
Statistical analysis
In the preference test with two wheels, the
number of revolutions made in each wheel
for the lO-day period was calculated, and
expressed as a percentage of the total number
made in both wheels. These values were
compared to 50% (no preference) using a one
sample two-tailed t-test.
Results
In Experiment I, a mesh-covered wheel was
overwhelmingly preferred to the standard
one with just rods for a base (97.7±0.9%,
L::J standard rods
Ell8l§l8mesh
_solid
~small
80
.2,.....
~.;60
~~
<uB
Q)<..,
$~
~
40
20
o
2
3
4
Experiment#
Fig 2 Mean ± SEM of preference for different wheel
types (% of total revolutions
on both wheels made in
10 days). Experiment 1: Standard rods vs plastic mesh.
Experiment 2: Plastic mesh vs solid floor. Experiment
3: Standard rods vs small wheels with rods. Experiment 4: Effect of experience in wheels with solid
floors
Laboratory Animals (2000) 34
316
Banjanin & Mrosovsky
# 414
WHEEL #1
~
CIl
>-
# 7565
WHEEL #1
.•.. MESH
~
.~.
...••
---.. ..•
..~..~
.•.. RODS
.•.. SOLID
-
.~
-
.•.. MESH
Q
'">-..:
Q
.•.. MESH
.•.. RODS
WHEEL #2
.•..
~
-~....-..
-~
----- .
MESH
.•.. SOLID
- ...
-
--------- ..
••...•..
--...
o
12
o
24h
24h
12
Fig 3 Actograms of representative
animals from Experiments 1 and 2. Upper arrowhead shows start of 10-day
data collection period, lower arrowhead shows when the wheel types were interchanged.
Open and closed bars
show times of light and dark. Wheel running on the y-axis is plotted in 15 quantiles, with the first including 180 revolutions
per 10 min bin, the second 81-160 etc
I
# 7575
WHEEL #1
# 7559
WHEEL #1
-----
--
..:
CIl
>-
-
..•..
'">-..: ..-
--
..•..STANDARQ..
Q
WHEEL #2
I.••
Q
•
-.
-
a
-
STANDARD
,.0.-
~--.......
..•..
SMALL
.
.•.. MESH
~~
-----
.•..
12
Actograms
of representative
24h
animals
~b·=~:
f-~-:-~_.
.•..
the mice had a mesh-covered wheel than the
standard one with rods (Fig 5). There was a
highly significant difference between the
groups in the total amount of running over
the IS-day period of recording (P = 0.0008,
two-tailed t-test).
Laboratory Animals (2000) 34
0
from Experiments
MESH
12
3 and 4. Conventions
-.•.•..••...
~-'~:
:~
~
o
- .•.- ~-
.....•...
....•.•....•.•.
WHEEL #2
.------------.-
--
--.L~
~L~_
1-
l- __
Fig 4
---.----
-..
~~
-
.•.. SOLID
.-~_-------.........---SOLID
.--:- .~
.
II
..I
24h
are the same as in Fig 3
Discussion
Wheels sometimes sold as mouse wheels
contain features that mice do not like: rod
flooring, and a relatively small size (13.0 cm
diameter). When given a choice in the
Running wheels for mice
317
U)
c
0 20000
S
~
~
Q) 10000
Q)
£
$:
10
5
15
Days
Fig 5 Mean± SEM wheel revolutions per day in two
different types of wheels (Experiment S). For the first
two points in the mesh group, data are based on n
values of 9 and 10 respectively, instead of 11, because
of problems with microswitches
present experiments, they showed preferences for larger wheels (17.5 em diameterl,
and wheels with plastic mesh flooring.
Choice tests are only one out of a number
of indicators of animal welfare (Dawkins
1980).Assessment of animal welfare is not
simple because sometimes inferences made
from different indicators do not agree (Mason
& MendI1993). Also, lack of a preferred
option is not always associated with poor
scores on other measures of welfare. For
instance, rats prefer solid to grid floors, but
no deleterious effects on food intake or
behaviour were detected in rats kept in grid
floors (Manser et al. 1995). Moreover, differences in ages, sex, or between individuals and
previous experience make blanket recommendations difficult (Mason & MendI1993).
Nevertheless, the preference for the plastic
mesh was present in mice of different ages
from different sources, and in both choice
and single wheel tests, suggesting it is relatively robust. However, until there has been
more extensive exploration, conclusions
should not be over-generalized, especially to
other strains of mice. With these cautions in
mind, it does seem reasonable to think that if
wheels are being used anyway in experiments
with mice, then larger ones with different
flooring from those commonly commercially
available should be employed. Not only do
mesh-covered wheels appear to be preferred
by the mice, but also when only one wheel is
available, as in Experiment 5, then the
amount of running is about 50% higher in
the mesh wheel compared with the standard
one (Fig 5). Promoting activity might be
helpful when wheel running is used as a
marker of circadian phase, especially in circumstances in which activity is depressed for
other reasons.
With respect to previous experience, the
present experiments showed that the preference for a mesh flooringover a solid plastic
base (Experiment 2, 71.6% of revolutions
made in the mesh-covered wheel) was absent
after experience with solid-covered wheels
only. It is not known whether previous
experience overcomes the stronger preference
(97.7%) for a mesh-covered wheel over a
wheel with rods. Mice are often conservative
in their behaviour (Kavanau & Brant 1965).
There are, however, some striking instances
of the effects of experience on wheel running.
For instance, mice (Peromyscus speciesl
given appropriate experience can come to
choose a square wheel, or one with small
hurdles, over a round wheel (Kavanau 1969).
However, an adaptation period to different
wheels lengthens experiments. It is simpler
to start with a wheel type that is preferred
initially by the animals. Therefore, we
recommend that use of wheels (17.5cm
diameterl with plastic mesh covers be
considered for experiments with mice.
Acknowledgments
We thank Peggy Salmon and
Uwe Redlin for help. Support came from the Medical
Research Council of Canada and the Natural Sciences
and Engineering Research Council of Canada.
The experiments were carried out in accordance
with the guidelines of the Canadian Council on
Animal Care.
References
Antoniadis E, McDonald RJ 11997)A conditioned
place preference paradigm reveals that wheel
running in the hamster is rewarding lAbstract
7268). Abstracts of the 27th Annual Meeting of the
Society of Neuroscience
Brant DH, Kavanau JL (1964) 'Unrewarded' exploration and learning of complex mazes by wild and
domestic mice. Nature 204; ?ft7-C)
Laboratory Animals (2000) 34
318
Collier CH, Johnson DF, CyBulski KA, McHale CA
(19901 Activity patterns in rats (Rattus norvegicusl
as a function of the cost of access to four resources.
Journal of Comparative Psychology 104, 53-65
Dawkins MS (1980) Animal Suffering, The Science of
Animal Welfare. London: Chapman and Hall
Iversen IH (1998) Simple and conditional visual
discrimination with wheel running as reinforcement in rats. Journal of the Experimental Analysis
of Behavior 70, 103-21
Kavanau JL 11969) Behavior of captive white-footed
mice. In: Naturalistic Viewpoints in Psychological
Research (Willems EP, Raush HL, edsl. New York:
Holt, Rinehart and Winston, pp 221-70
Kavanau JL, Brant DH (1965) Wheel-running preferences of Peromyscus. Nature 208, 597-8
Manser CE, Morris TH, Broom DM 11995)An
investigation into the effects of solid or grid cage
flooring on the welfare of laboratory rats. Laboratory Animals 29, 353-63
Laboratory Animals (2000) 34
Banjanin & Mrosovsky
Mason C, Mendl M (1993) Why is there no simple
way of measuring animal welfare? Animal Welfare
2, 301-19
Moore-Ede Me, Sulzman FM, Fuller CA 119821The
Clocks That Time Us; Physiology of the Circadian
Timing System. Cambridge: Harvard University
Press
Mrosovsky N, Salmon PA, Vrang N (1998) Revolutionary science: an improved running wheel
for hamsters. Chronobiology International 15,
147-58
Sherwin CM (1998al Voluntary wheel running: a
review and novel interpretation. Animal Behaviour
56, 11-27
Sherwin CM 11998b)The use and perceived importance of three resources which provide caged
laboratory mice the opportunity for extended
locomotion. Applied Animal Behaviour Science
55,353-67