Ann Baldwin, Ph.D. Research Professor of Physiology, University of Arizona

According to the U.S. Department of Agriculture (USDA), rodents comprise approximately 90% of all animals used in research today. In order to ensure the validity of data collected from these animals, it is crucial that they are kept under conditions that do not cause them stress. Stress adversely affects every physiological system, thereby introducing a confounding variable into experimental designs.

One potential source of stress is the animal housing environment. Chronic environmental stress due to lack of enrichment imposes a host of adverse physiological consequences on rodents, including an increase in corticosterone levels and the development of repetitive behaviors (e.g. excessive grooming, digging, rearing, yawning, and fighting/biting). However, research rodents are often housed in small plastic cages that lack items common to their natural environment and limit their opportunities to perform their natural behaviors.

In this article, Dr. Baldwin describes two studies, one on rats (Brauner et al, in press) and on on mice (Cudilo et al, 2007), demonstrating that lack of enrichment can totally change physiological and pathological measures, leading to erroneous conclusions.

Heart Rate Variability in Rats
Heart rate variability (HRV) is how the heart rate varies with time resulting from a variety of factors including neural input from the parasympathetic and sympathetic nervous systems. Sympathetic activation tends to produce low frequency (LF: 0.05 0.15 Hz) oscillations in heart rate, whereas parasympathetic activation produces higher frequency (HF: 0.15-0.40 Hz) oscillations. By comparing the relative ‘powers’ of the two frequency ranges of oscillations, LF/HF, it is possible to determine an animal’s sympathovagal balance. A significant increase in LF/HF is representative of increased stress.

Heart rate variability is commonly used to predict clinical outcomes in trials involving treatment of heart disease in humans. This is relevant to rodents because rats are considered a good model for cardiovascular disease. It is important that potential factors that can confound HRV in rats used for research are identified in order to avoid imprecise results in drug trials. Enrichment is not the only component of the cage that could impact a rodent’s stress. The size of the cage relative to the number of animals in that cage could also significantly impact stress. The goal of this study was to evaluate the effects of cage size and enrichment on LF/HF of rats housed in one of two standard sizes of rodent cages and provided with or without two enrichment items (tube and shelf).

Before the experiment, the 10 rats were housed in pairs in large, enriched cages because our previous preliminary studies (Baldwin et al, 2005) showed that rats housed in large, enriched cages demonstrated less aggressive nocturnal behavior than those housed in small, un-enriched cages. The cages were located in a university animal facility with a 12 hour light-dark cycle (lights on at 6 AM and off at 6 PM). One of each pair of rats was pre-implanted with a telemetric transducer (C50 PXT, Data Sciences International, St. Paul, MN) to allow remote measurement of ECG from which LF/HF was derived.

At the start of the experiment, the rats were housed in the small un-enriched cage (SU) and (after the first 3 week assessment) were randomly assigned to each of the other three cage conditions [small enriched (SE), large un-enriched (LU), and large enriched (LE)] until they had experienced each condition once. All of the cages contained a layer of pine shavings as bedding. Large cages provided a floor area of 3.5 cm2 per gram weight (350 g rats) or 4.0 cm2/g (500 g rats), and small cages provided the rats with a floor area of 2.5 cm2 per gram weight. The enrichment items consisted of a polyvinyl chloride tube and a wire mesh shelf to increase the complexity of the cage while stimulating the rodent’s natural species-specific behaviors (nesting behaviors and subordinate rat escape behaviors). During the first week of each cage condition, the rats acclimated to their new surroundings. For the next two weeks ECG data were collected and the rats were videotaped for behavioral analysis twice a day (8 AM and 8 PM) for 10 minutes, three days a week. Rat behaviors involving activity were classified from video recordings by means of an established Rat Ethogram and the percentage of total time (AM and PM) each rat spent performing active behaviors was evaluated.

There was no difference in LF/HF between the four cage conditions when considered independent of sleep/wake cycle but LF/HF increased when the rats were awake and active (p<0.05, F=32.3). Since the HF component (primarily parasympathetic nervous activity) was not different, regardless of cage condition or time of day, the increase in LF/HF ratio reflects an increase in sympathetic nervous activity (SNA). The amount of time spent in the active state increased during the evening (p<0.05, F=80.47). The increase in LF/ HF seen when the rats were awake compared to asleep was driven by the un-enriched cage condition (p<0.05, F=5.63) as no significant change in LF/HF (PM vs. AM) was observed in the enriched environment. On the other hand, the differences in activity levels observed between AM and PM were seen in both enriched and un-enriched conditions (p<0.05). In summary, the data suggest that enrichment significantly reduces the difference in LF/HF experienced by the rats throughout the sleep/wake cycle in the un-enriched cage condition and that this effect cannot be explained by a reduced variation in activity levels.

The increases in LF/HF and activity seen when the rats were awake occurred for both the small and large cage conditions. Thus an increase in cage size above the recommended minimum, regardless of the presence or absence of enrichment, was not sufficient to reduce the difference in LF/HF experienced by the rats throughout the sleep/wake cycle.

What does this study tell us?
Addition of enrichment, regardless of cage size, significantly reduced the apparent diurnal rhythm in LF/ HF. This finding is not surprising because HRV is a very sensitive physiological measure that is affected by emotions. Interestingly, another study showed that when miniature swine were housed together in pairs instead of in isolation, the diurnal rhythm of LF/HF also disappeared (Kuwahara et al, 2004). These results suggest that the apparent diurnal rhythm of LF/HF is an artifact in caged animals, only seen when animals are prevented from performing species-specific behaviors.

Arterial Pathology in Knockout Mice
Fibulin proteins play an important role in maintaining the mechanical properties of artery walls. Fibulin-4 is an extracellular matrix protein expressed by vascular smooth muscle cells and is essential for maintaining arterial integrity. Fibulin-4-/-mice, in which both fibulin-4 genes are knocked out, die just before birth due to arterial hemorrhage, but fibulin-4+/-mice, in which only one gene is knocked out, appear to be outwardly normal. A colleague of Dr. Baldwin’s, Dr. Lihua Marmostein, asked Dr. Baldwin whether she would perform experiments to determine if the fibulin-4+/-mice showed normal arterial structure on a microscopic scale.

Dr. Baldwin performed preliminary experiments on fibulin4+/-mice housed in the usual way (four mice per cage in standard cages (26 cm (length) x 16 cm (width) x 12 cm (height)) containing bedding but no enrichment). Electron microscopy showed localized regions of disorganized extracellular matrix and collagen fibers or ‘gaps’ between some of the medial smooth muscle cells in the mouse aortas. Similar experiments performed on wild-type mice with both fibulin-4 genes intact showed that the smooth muscle cells of the aorta were closely connected to each other and the media was more compact. The number of gaps per square mm was more than ten times greater for fibulin-4+/-mice (172 ± 43 (SEM)) than for wild-type mice (15 ± 8) (p <0.01, n=8).
Dr. Baldwin was rather disturbed by the sterile, unstimulating conditions in which the mice were housed and decided to repeat the experiments on mice housed, two per cage, in larger cages (33 cm (length), 25 cm (width) x 25 cm (height)) that contained a shelf, ladder, exercise wheel and a plastic tube. In the enriched cages where the mice could run, climb and nestle in the tunnel, the number of gaps in the fibulin-4+/-mice (35±12) was reduced almost to wild-type amounts and was significantly lower than for fibulin-4+/-mice in the standard cages (p<0.05, n=8).

What does this study tell us?
Dr. Baldwin’s team demonstrates for the first time a connection between a genetically determined, vascular disease and environment affecting the degree of manifestation of disease symptoms. The study also sheds light on the fact that scientists should pay careful attention to housing conditions and bear in mind that differences in lifestyle could account for varying results. Thus, research findings assumed to be attributed to genetic differences might be interpreted incorrectly, neglecting the role of environmental factors.

Anna E. Brauner, M.S., David T. Kurjiaka, Ph.D. and Angela Ibragimov, B.S. were on the team for the rat experiments. Elizabeth Cudilo, Hamda Al Naemi and Lihua Marmostein were on the team for the mouse experiments.

Baldwin AL, C Vincifora, T Burke & M Gritzuk: Effect of rodent housing on behavior and microvascular leakage. The FASEB Journal, 2005, 19,5, #691.11, A1263.
Brauner AE, Kurjiaka DT, Ibragimov A & Baldwin AL. Impact of cage size and enrichment , (tube and shelf) on heart rate variability in rats.
Scandinavian Journal of Laboratory Animal Science, 37(2), (in press).
Cudilo E, Al Naemi H, Marmorstein L and Baldwin AL: Knockout mice: is it just genetics?
Effects of enrichment on fibulin-4+/- mice.PLos ONE, 2007, 2(2): e229.
Kuwahara M, Y Tsujino, H Tsubone, E Kumagai & M Tanigawa: Effects of pair housing on diurnal rhythms of heart rate and heart rate variability in miniature swine. Exp. Anim., 2004, 53, 4, 303-309.

Issue 5 October 2010

View this issue in flipbook format

%d bloggers like this: