Online Issues
<< All Back-issues
<< This Issue's Table of Contents
ILAR Journal V39(1) 1998
Animal Well-being: Immune Function, Behavior, Morphology, and Psychoneuroimmunology
Steven Fullwood, Tiffanie A. Hicks, Jack C. Brown, Reid L. Norman, and John J. McGlone
| Steven Fullwood, B.S., Tiffanie A. Hicks, M.S., and John J. McGlone, Ph.D., are Graduate Assistant, Research Associate, and Professor, respectively, in the Department of Animal Science, Texas Tech University, Lubbock. Texas. Jack C. Brown, Ph.D., is Professor in the Department of Microbiology, University of Kansas, Lawrence, Kansas. Reid L. Norman, Ph.D., and Dr. McGlone are Professors in the Department of Cell Biology & Biochemistry, Texas Tech University Health Sciences Center, Lubbock, Texas. |
INTRODUCTION
The welfare of laboratory animals remains a concern to the general public and to laboratory animal users. Furthermore, scientists hope to use animal housing systems that do not impose any unforeseen stress. Floor space is provided for laboratory animals based on concerns about animal well-being, animal performance, and economics--rather than by a rational examination of scientific data.
An examination of the scientific literature that forms the basis of current recommendations for space for laboratory mice reveals that the floor space recommendations are not based on controlled investigations. Indeed, we found a paucity of information on space needs for mice.
Many of the early studies of "crowding" in mice were conducted with the idea of studying the physiological basis of stress but not of establishing a minimum space requirement. Investigators commonly induced crowding and used a speculative control treatment that provided enough space to not cause the stress response to crowding. In other instances, many studies confounded space per animal with group size (mice per cage). For example, Christian (1955) used cages of 935.5 cm2 to house laboratory mice in groups of 1, 4, 8, 16, and 32 and wild mice in groups of 1, 3, 4, 6, 8, 9, and 17 mice per cage. The crowding effect was composed of the confounded effects of less floor space per mouse plus the combined group-size effects that the increasing social pressures precipitated to create the experience of crowding. Many authors (Barnard and others 1994; Christian and others 1961; Gammallo and others 1986; Hull and others 1976; Jean-Faucher and others 1981; Ortiz and others 1985; Peng and others 1989) directly confounded the effects of floor space per mouse and group size in similar models to study crowding. Defining the floor space that induced the stress response was not an objective of the early work; instead, mice were added to cages until the desired response was achieved. The investigators were concerned more with the physiological mechanisms involved during crowding than in recommending a floor area that was not stressful to mice.
We know that many factors will potentially interact with the establishment of minimum floor space needs for mice and other animals. We are especially concerned about important effects of cage type, floor surface (bedded or wire), genetic line, gender, and cage changing frequency. Thus, in this study, we sought to investigate the floor space needs of mice of only 1 strain (C57BL/6) of male mice in 1 cage type (solid floor with bedding). To assess mouse welfare, we used measures of growth, survival, adrenal hormones, and immune status. We also recorded extensive video tapes so behavioral measures may be collected in the future. The specific objective of this study was to determine the growth and endocrine and immunological responses of C57BL/6 male mice to floor spaces of 5, 10, 15, and 20 in2 (32.2, 64.5, 96.8, and 129 cm2) per mouse (the NRC 1996 recommendation was 15 in2 [or 96.8 cm2] per mouse).
METHODS
General
Mice (C57BL/6) were purchased (Charles River Laboratories, Raleigh, North Carolina) at 4 to 5 wk of age. The mice were given 1-wk acclimation in group housing on bedding (Heat Treated Sani-Chips, P. J. Murphy Forest Products, Montville, New Jersey) in shoe boxes and were provided food (Harlan TEKLAD Laboratory Diet 22/5 Rodent Diet W 8640) and water ad libitum. Standard hanging cages (L 9 7/ 8 x W 7 1/8 x H 7 in [L 24.45 x W 18.09 x H 17.78 cm]) (Shore Line, Kansas City, Missouri) were modified by adding a galvanized steel floor plus a wall to impose different floor area while maintaining a constant group size of 3 mice per cage. Thus, the following floor areas were provided for each treatment of 5, 10, 15, and 20 in2 per mouse (32.2, 64.5, 96.8, and 129 cm2), respectively: 7.125 x 2.12 in (18.10 x 5.34 cm); 7.125 x 4.21 in (18.10 x 10.69 cm); 7.125 x 6.34 in (18.10 x 16.04 cm); and 7.125 x 8.42 in (18.10 x 21.38 cm). The wall was placed parallel to the back wall of the cage to not limit the area for feed and water. The entire floor area was covered in galvanized steel with a 1-in (2.54-cm) lip on the wire mesh front for the addition of bedding. Any food loss occurred through the wire mesh front through which the mice pushed out the bedding and food often. The floor of galvanized steel was installed so that bedding could be added. Animals were placed in treatments of 5, 10, 15 (NRC 1996 recommendation), or 20 in2 (32.2, 64.5, 96.8, or 129 cm2) per mouse. Social groups of 3 mice per treatment were used, and the mice were treated for 5 wk. High and low air temperature and relative humidity were recorded daily for the room. Mortality and general health data were collected daily. The bedding in each cage was checked daily and added as needed. The bedding in the smaller floor areas was changed more often due to the dampness of the bedding. The mice caused the increased dampness by bumping or deliberately moving the dripper. Once a week, when the cages were exchanged with clean cages that had fresh bedding, each mouse was weighed on a standard triple beam balance. Also at that time the feed left in the feeder that hangs off the front of the cage was recorded using a triple beam balance, the water left in the bottles was measured using a graduated cylinder, and new feed and fresh water were added. Feed and water wastage was not measured directly. Weekly data were collected for 5 wk.
Immune Measures
At the end of the growth period, mice were anesthetized with CO2, and 1 to 2 mL of blood were collected over sodium heparin by severing the brachial artery. Adrenal glands were weighed, and the spleen was removed from each mouse. Mice blood or cells were assayed for natural killer (NK1) cell activity, mitogen-induced lymphocyte proliferation, and differential counts. Blood smears made using whole blood were fixed in methanol and stained with Hemo-3 (Biochemical Sciences, Bridgeport, New Jersey) for differential white cell counts; 100 cells were counted per slide. At 1:500 dilution, the number of white blood cells and the hemoglobin concentration were determined using a Coulter cell counter (Coulter Electronics, Hialeah, Florida). The NK assay was performed by published techniques (Lumpkin and McGlone 1992). Briefly, spleens were removed from the mice and teased apart to release the effector cells. The targets for the assays YAC-1 cells, a murine chronic mylogenous leukemia cell line (American Type Culture Collection, Rockville Pike, Maryland), were labeled with 51Cr. Samples were run at effector:target ratios of 6.25:1, 12.5:1, 25:1, and 50:1; 104 target cells were added to each well. Percentage of cytotoxicity was calculated as previously described (Lumpkin and McGlone 1992). The mitogen-induced lymphocyte proliferation assay was used to determine proliferation of lymphocytes according to an established method (Morrow-Tesch and others 1994). In short, the washed mononuclear cells obtained from the spleens were counted on the Coulter counter and adjusted to a concentration of 5 x 105 cells/mL. Each sample (5 x 105 cells/mL) was added in triplicate to the wells of microtiter plates containing 100 mL of each mitogen of phytohemmaglutinin (PHA1) (Sigma Chemical, St. Louis, Missouri) in RPMI (200 mL of total volume) and lipopolysaccharide (Sigma). The plates were incubated for 48 hr and then pulsed with 50 mL of tritiated thymidine (20 mCi/mL). The plates were harvested 24 hr later using a cell harvester (Brandel Inc., Gaithersburg, Maryland), and the filter disc was put into scintillation vials and counted.
Endocrine Measures
In addition to the paired adrenal weights, plasma was frozen for determination of plasma corticosterone. We assayed for cortisol as well, but its concentration was nondetectable. The assay was performed from a kit supplied by Diagnostic Products Corporation (Los Angeles, California). Briefly, serum was placed in an antibody-coated tube to which the sample and then I125-labeled cortisol had been added. After an incubation, the tube was thoroughly decanted and counted on a gamma counter. The plasma corticosterone assay was performed in a single assay with an intraassay coefficient of variation of 3.6%.
Statistical Analyses
Measures of body weight changes, feed and water disappearance, and immune and adrenal measures were subjected to analysis of variance using the General Linear Models procedure of SAS (1997). All reported measurements were taken on all surviving mice. The experimental design was a randomized complete block, with 4 treatments (each space allowance) and 6 blocks. A block was a set of each treatment with 3 mice per cage; each block contained 12 mice (72 mice in total). A block represented 12 mice that were housed 3 per cage in each of the 4 treatments. Thus, there were 6 blocks comprising a total of 72 mice. The 6 blocks represented 6 exact replications of the study. The experimental unit was the cage of 3 mice. Thus, the statistical model had 24 observations or cages.
Data were analyzed separately for each week to learn weight, mitogen dose for lymphocyte proliferation, and effector:target ratio for NK assay. Treatment means were separated with the Predicted Difference test within the General Linear Models procedure (SAS 1997). In addition to the analysis of variance described above, linear regression was used to estimate possible linear, quadratic, or cubic relationships between floor space and the measures above. Mortality data were subjected to chi-square analysis.
RESULTS
Mouse body weight changes over time were only marginally influenced by floor space treatments (Table 1 ). Mice in each treatment gained weight. Limited floor space did not slow weight gain over the 5-wk evaluation period. However, water and feed disappearance were influenced (P < 0.01) by floor space treatments (Table 1); with smaller floor spaces, mice appeared to use more feed and water. We attributed the increased feed and water use to greater feed wastage. We commonly observed mice dropping feed, water, and bedding into the waste pans below the cages. Mice at 5 in2 (32 cm2) had significantly greater feed and water disappearance than mice at greater space allowances during the last week of the study. Mice at 10 in2 per mouse had elevated feed and water disappearance compared with mice given greater floor space during the last week of data collection. However, feed and water disappeared less for mice at 10 in2 (64 cm2) than for mice with floor space of 5 in2 (32 cm2) per mouse. The average room temperature was 71.4°F (21.9°C), and the average humidity was 62.3% for the entire time of the study.
Immune data were collected for the mice after 5 wk in each environment (Table 2). The lymphocyte proliferation assay showed an effect for the mitogen at a concentration of 0.2 g/mL of PHA (P < 0.05) and for 0 g/mL of PHA (the control) (P < 0.05). The mice in 5 in2 per mouse space had a significantly greater lymphocyte blastogenesis than all other treatments. In addition, the NK showed an effect at an effector:target ratio of 12.5. In this case, the limited space of 10 in2 (32 cm2) per mouse caused a significant increase in NK activity compared with mice given 15 or 20 in2 (97 or 129 cm2) per mouse.
Adrenal gland weights were not significantly influenced by space allowance by analysis of variance; however, we observed a linear increase in adrenal gland weight with smaller cage sizes (Figure 1), which followed the results of plasma corticosterone concentrations. Plasma corticosterone concentrations were influenced (P < 0.05) by space allowances: Mice given smaller spaces had elevated corticosterone concentrations (Figure 2). Mortality was significantly less in the crowded (5 in2 [32 cm2] per mouse) than in all other treatments (P < 0.01) We observed greater mortality in the mice provided more space (Figure 3). The cause of death for each mouse was determined by the university veterinarian to be due to bite and attack wounds. The majority of the mice were bitten on the spine, resulting in paralysis. When the mice were found injured, they were immediately and humanely euthanized. Other deaths occurred due to starvation brought on by the competition over food. Animals observed dying were euthanized.
DISCUSSION
Among farm animals, limited space reduces weight gains (Brumm and Dahlquist 1995; Brumm and NCR-89 Committee on Management of Swine 1996; McGlone and Newby 1994). In spite of providing quite small spaces for mice, their weight gains and body weights did not differ after 5 wk of apparent crowding. Several explanations are possible: (1) Our mice, even at 5 in2 per mouse, were not crowded; (2) all treatment space allowances were crowded: or (3) mice do not respond as other animals by limiting weight gain when they are in very small spaces. Indeed, the argument could be made, based on body weight changes alone, that mice are not particularly crowded at even 5 in2 per mouse. However, other measures would not support this conclusion.
Feed and water disappearances were greatly influenced by the amount of floor space provided (Table 1). Very limited floor spaces (5 in2 per mouse) significantly increased feed and water use. Because we observed increased feed and water wastage among the mice with limited floor space, we speculate that limited floor space leads to feed and water wastage. This finding could have an impact both on studies seeking to measure feed and water intake and on the economics of mouse housing in that reduced space will increase feed cost. The increased water use will require more frequent cage changes due to wet bedding.
Immune measures were also influenced by floor space allowance. For lymphocyte proliferation in response to PHA mitogen, reduced floor space tended to increase or immunostimulate splenic lymphocytes. This effect was not observed for the lipopolysaccharide mitogen. Thus, while mouse T-helper cells may be stimulated by reduced spaces, the range of spaces evaluated did not influence B-cell proliferation. We believe the T-helper cells of mice kept at the NRC 1996 recommended space allowance ( 15 in2 or 97 cm2 per mouse) are less responsive to mitogen than mice allowed less space. Alternatively, the apparent immunostimulation could be due to a change in numbers of functionally responsive cells. For example, elevated glucocorticoids could lyse some subsets of lymphocytes, resulting in immunomodulation. To investigate this hypothesis, one should also study lymphocyte subcell distributions in putative stressful environments.
The maximum NK activity of these mice likewise favored smaller spaces. At 10 in2 per mouse, their NK activity was greater (P < 0.05) than that of mice allowed the NRC 1996 recommended space per mouse.
For both lymphocyte proliferation under a T-cell mitogen and NK cell activity, use of the NRC 1996 space allowances caused an apparent immunosuppression. The immunosuppression of mice at NRC 1996 space allowances were of the same magnitude as rodents experiencing many stressors such as 2-DG metabolic, burn, or restraint stress (Miller and others 1994; Penturf and others 1996; Anisman and others 1997; Chou and others 1997).
The interpretation of plasma glucocorticoid concentrations and adrenal weights was in contrast to the interpretation of immune measures. Adrenal responses were greater when floor space was smaller (Figures 1 and 2). The observation that adrenal weights and glucocorticoid concentrations were influenced similarly by the spaces provided lends credibility and consistency to these measures.
Not all studies of crowding show an increase in plasma glucocorticoid concentrations or adrenal weight, especially when group size is increased as floor space decreases. Ortiz and others (1985) housed 4, 8, or 24 mice in 529 cm2
At 24 mice per cage, their OF1 male mice had only 22 cm2 per mouse (30% less space per mouse than our most limited space allowance). At this time we cannot say whether our different results are due to (1) mouse strain differences, (2) a protective effect of large group sizes during limited spaces, or (3) some other factor(s). However, we do know from models developed with pigs that as group size increases, the per animal space needs decrease (McGlone and Newby 1994). We also know that mouse strains vary in their physiological responses to quantity and quality of space (Peters and Festing 1990).
Although elevated glucocorticoid activity is often seen as a sign of stress, the immune measures and the endocrine measures lead us to opposite conclusions. The mortality data--the ultimate measure of an environment's fitness--reflected the immune data almost exactly. The observed mouse mortality was greater when the space provided was greater: Repeated defeats of submissive mice resulted in mortality, whereas mice that had sufficient room to maneuver were bitten, had hair pulled out, and had whiskers barbered away.
Scientists who study animal well-being conclude that multiple measures are needed to decide about adequate animal welfare (Curtis and others 1997). Based on the present investigation, one would have to conclude that limited space (relative to NRC 1996 recommendations) promotes better mouse welfare. Furthermore, precisely following the NRC 1996 recommendations--using social housing and 15 in2 per mouse--would reduce the welfare (measured by immune and mortality measures) of mice. Floor space of 5 or 10 in2 per mouse promoted no lower and in some cases greater NK activity or PHA-stimulated lymphocyte proliferation and lower mortality than 15 in2 per mouse. Indeed, for male mice of this strain, more space caused a brutal and deadly situation for the submissive cage mates. In spite of adrenal enlargement, smaller spaces promoted better overall immune activity and lower mortality.
Clearly more data are needed to understand the biology of common laboratory species in common housing systems. Human intuition and professional judgment do not necessarily lead to the same conclusion about housing needs as physiological data from controlled studies. More aspects of physiology and behavior must be studied before sound husbandry decisions can be made.
SUMMARY
Measures of performance, mortality, adrenal weights, plasma glucocorticoid concentration, and selected immune measures were collected in an attempt to define space needs of laboratory mice. Six replications of 3 C57BL/6 male mice per cage were examined while housed on bedding at 5, 10, 15, or 20 in2 (32.2, 64.5, 96.8 or 129 cm2) per mouse. Body weights were not influenced by treatment; however, mice in smaller spaces (5 in2 per mouse) consumed or wasted more feed and water than mice given greater space allowances. Mice given the least amount of space (5 in2 per mouse) had greater lymphocyte proliferation in response to the T-cell mitogen PHA than mice given more space. Mice provided 10 in2 per mouse had greater NK cytotoxicity than mice given greater or less space. Mouse mortality was greater as more space was provided. In contrast, adrenal weights and plasma glucocorticoid concentrations were progressively greater with lower space allowances. The NRC 1996 recommendation of 15 in2 per mouse, for this strain and sex of mice, would result in greater mortality and reduced activity of some immune measures. Socially housed male C57BL/6 mice will benefit from less space than recommended by the NRC in 1996.
1Abbreviations used in this paper: NK, natural killer; PHA, phytohemmaglutinin.
ACKNOWLEDGMENTS
This work was supported by the State of Texas. This work (protocol no. 97665) was approved June 25, 1997, by the Texas Tech University Institutional Animal Care and Use Committee.
REFERENCES
Anisman H, Lu ZW , Song C, Kent P, McIntyre DC, Merali Z. t997. Influence of psychogenic and neurogenic stressors on endocrine and immune activity: Differential effects in fast and slow seizing rat strains. Brain Behav Immun 11:63-74.
Barnard CJ, Behnke JM, Sewell J. 1994. Social behavior and susceptibility to infection in house mice (Mus muscularis): Effects of group size, aggressive behavior and status-related hormonal responses prior to infection on resistance to Babesia microti. Parasitology 108:487-496.
Brumm M, Dalquist J. 1995. Nursery and Growing-Finishing Space Interactions. Nebraska Swine Report 44-45.
Brumm MC and NCR-89 Committee on Management of Swine. 1996. Effect of space allowances in barrow performance to 136 kilograms body weight. J Anim Sci 74:745-749.
Christian JJ. 1955. Effect of population size on the adrenal glands and reproductive organs of male mice in populations of fixed size. Am J Physiol 182:292-300.
Christian JJ. 1961. Phenomena associated with population density. Proc Natl Acad Sci U S A 47:428-449.
Chou SH, Ljiljana DK, Cunnick JE. 1997. Evidence for the involvement of catecholamines in the 2-DG-induced immunomodulatory eflk2cts in spleen. Brain Behav Immun I 1:79-93.
Curtis SE, editor. 1997. The Well-Being of Agricultural Animals. Ames IA: Council for Agricultural Science and Technology.
Gammallo A, Villanava A, Trancho G, Fraile A. 1986. Stress adaption and adrenal activity in isolated and crowded rats. Physiol Behav 36:221-317.
Hull EM, Kastaniotis C, L'Hommedieu G, Franz J. 1976. Environmental enrichment and crowding: Behavioral and hormonal effects. Physiol Behav 17:735-741.
Jean-Faucher Ch, Berger M, Turckheim M, Veyssiere G, Jean CL. 1981. Effects of dense housing on the growth of reproductive organs, plasma testosterone levels and fertility of male mice. J Endocrinol 90:397-402.
Lumpkin EA, McGlone JJ. 1992. A 51Cr release assay for determination of natural killer cell cytotoxicity. J Nutr Immunol 1:167-74.
McGlone JJ, Newby BE. 1994. Space requirements for finishing pigs in confinement: Behavior and performance while group size and space vary. Appl Anim Behav Sci 39:331-338.
Miller ES, Klinger JC, Akin C, Koebel DA, Sonnenfeld G. 1994. Inhibition of murine splenic T lymphocyte proliferation by 2-deoxy-D-glucose-induced metabolic stress. J Neuroimmunol 52:165-173.
Morrow-Tesch JL, McGlone JJ, Salak-Johnson JL. 1994. Heat and social stress effects on pig immune measures. J Anim Sci 72:2599-2609.
NRC [National Research Council]. 1996. Guide for the Care and Use of Laboratory Animals. Washington DC: National Academy Press.
Ortiz R, Arimario A, Castellanos JM, Balasch J. 1985. Post-weaning crowding induces corticoadrenal hyperactivity in male mice. Physiol Behav 34:857-860.
Peng X, Lang CM, Drozdowicz CK, Ohlsson-Wilhelm BM. 1989. Effect of cage population density on plasma corticosterone and peripheral lymphocyte populations. Lab Anim 23:302-306.
Penturf M, McGlone JJ, Griswold JA. 1996. Modulation of immune response in thermal injury by essential fatty acid diet. J Burn Care Rehabil 17(5):464-470.
Peters A, Festing M. 1990. Population density and growth rate in laboratory mice. Lab Anim 24:273-279.
SAS Institute, Inc. 1997. SAS user's guide: Statistics, version 6.08, KP Ingraham, editor. Cary NC: SAS Institute, Inc.
TABLE 1 Least squares means for weight, feed intake, and water intake over 5 wk
| Treatment space per mouse | ||||||
| Measure per mouse in2 cm2 | 5 32.2 | 10 64.5 | 15 96.8 | 20 129.0 | SE | Treatment P value |
| N = 6 replicates of 3 mice per mean | ||||||
| Body weight (g) | ||||||
| Time 0 | 22.7 | 22.4 | 21.9 | 21.8 | .29 | 0.15 |
| Week 1 | 21.5 | 22.5 | 21.4 | 21.7 | .44 | 0.39 |
| Week 2 | 22.7 | 24.4 | 21.4 | 23.2 | .81 | 0.13 |
| Week 3 | 24.4 | 25.9 | 24.0 | 24.7 | .60 | 0.20 |
| Week 4 | 25.8 | 26.5 | 24.2 | 25.1 | .52 | 0.07 |
| Week 5 | 26.4 | 26.7 | 26.4 | 25.5 | .51 | 0.41 |
| Sacrifice | 26.6 | 26.8 | 26.0 | 26.0 | .59 | 0.28 |
| Feed intake (g) | ||||||
| Week 1 | 87.3 | 64.7 | 68.5 | 55.3 | 9.58 | 0.15 |
| Week 2 | 94.3 | 70.9 | 57.7 | 66.3 | 8.68 | 0.05 |
| Week 3 | 94.1 | 86.3 | 81.0 | 85.8 | 4.87 | 0.31 |
| Week 4 | 103.4a | 85.8b | 75.1b | 76.1b | 5.52 | 0.008 |
| Week 5 | 108.1a | 92.0b | 78.1c | 84.6b | 3.98 | 0.0009 |
| Water intake (mL) | ||||||
| Week 1 | 193.5 | 90.0 | 91.3 | 104.5 | 33.37 | 0.13 |
| Week 2 | 143.6 | 115.9 | 194.2 | 142.3 | 33.47 | 0.42 |
| Week 3 | 145.2 | 136.5 | 122.7 | 130.3 | 23.84 | 0.92 |
| Week 4 | 144.2 | 186.0 | 100.8 | 137.8 | 28.16 | 0.23 |
| Week 5 | 160.8a | 128.8b | 100.8c | 105.6c | 7.99 | 0.0004 |
SE, pooled standard error.
P value, treatment effect for the analysis of variance.
TABLE 2 Least squares means for immune measures
| Treatment space per mouse | ||||||
| Measure per mouse in2 cm2 | 5 32.2 | 10 64.5 | 15 96.8 | 20 129.0 | SE | Treatment P value |
| Number | 18 | 15 | 9 | 12 | ||
| Whole blood measures | ||||||
| WBC x 106 cells/mL | 12.3 | 11.4 | 16.4 | 10.6 | 4.82 | 0.88 |
| Hemoglobin g/dL | 11.5 | 10.6 | 12.2 | 10.2 | .98 | 0.55 |
| Monocyte x 106 cells/mL | 7.17 | 9.13 | 8.00 | 11.0 | 2.21 | 0.56 |
| Neutrophil x 106 cells/mL | 9.76 | 5.06 | 9.25 | 11.0 | 2.26 | 0.42 |
| Lymphocyte x 106 cells/mL | 83.1 | 86.7 | 82.8 | 78.0 | 3.52 | 0.44 |
| Netrophil:lymphocyte ratio | 9.73 | 12.0 | 10.8 | 16.1 | 3.51 | 0.67 |
| Lymphocyte proliferation, dose, mitogen | ||||||
| LTA 20 mg/mL PHA | 84,457 | 58,776 | 72,422 | 61,186 | 13,855 | 0.60 |
| LTA 0.2 mg/mL PHA | 87,974a | 56,237b | 63,904b | 63,993b | 7232 | 0.01 |
| LTA 0 mg/mL PHA | 59,814a | 46,347b | 49,904a,b | 42,917b | 4,109b | 0.02 |
| LTA 20 mg/mL LPS | 81,910 | 81,152 | 78,821 | 95,267 | 7678 | 0.46 |
| LTA 0.2 mg/mL LPS | 73,722 | 77,932 | 68,278 | 75,647 | 5801 | 0.80 |
| LTA 20 mg/mL LPS | 76,423 | 48,869 | 67,233 | 52,395 | 13,864 | 0.38 |
| NK cell activity | ||||||
| NK % cytotoxicity, E:T=50 | 7.65 | 6.58 | 4.48 | 4.58 | 1.44 | 0.23 |
| NK % cytotoxicity, E:T=25 | 3.87 | 3.73 | 3.25 | 3.10 | 1.05 | 0.94 |
| NK % cytotoxicity, E:T=12.5 | 2.23b | 6.00a | 1.54b | 2.04b | 1.13 | 0.02 |
| NK % cytotoxicity, E:T=6.25 | 1.84 | 2.68 | 1.34 | 1.11 | .72 | 0.40 |
P value, treatment effect for the analysis of variance.
E:T, effector:target ratio; LPS, lipopolysaccharide mitogen; LTA, lymphocyte transformation assay; NK, natural killer; PHA, phytohemmaglutinin mitogen; SE, pooled standard error.
FIGURE 1 Adrenal gland weights (paired weight, mg) for mice in each space allowance, with cage size in cm2. The regression equation (Y = 7.125 + .0774X) had an r = 0.92. The analysis of variance evaluating treatment effect was not significant (treatment effect P = -.68).
FIGURE 2 Plasma corticosterone concentrations in ng/mL, with cage size in cm2. Plasma corticosterone was influenced by floor space treatment (P=0.05), and the linear effect explained the relationship between floor space and plasma corticosterone concentration (Y=381.7 + 12.06X, r=0.91).
FIGURE 3 Mortality (%) of mice on the study, with cage size in cm2. Data (analyzed by chi-square) showed a significant (P<0.01) effect of floor space on the rate of death of male C57BL/6 mice.
Copyright © 2008. National Academy of Sciences.
All rights reserved.
500 Fifth St. N.W., Washington, D.C. 20001.
Terms of Use and Privacy Statement