167

Differential Effects of Exercise and Housing Condition on Murine Natural Killer Cell Activity and Tumor Growth L. Hoffman-Goetz"2 , B. MacNeil2, Y. Arumugam' , J. Randall Simpson' 'Department of Health Studies and 2Department of Kinesiology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1

There was a significant effect of physical activity (p < 0.0 19) but not of housing on splenic NK cytotoxicity

Activity and Tumor Growth. mt J Sports Med, Vol 13, No 2,

against tumor targets in vitro. When the data were analyzed by presence or absence of lung metastases, only those animals without visible lung tumors had significantly higher NK activity as a function of exercise relative to sedentary

ppl67—l7l,l992.

controls. There were no significant differences in the

L. Hoffman-Goetz, B. MacNeil, Y. Arumugam and J. Randall Simpson, Differential Effects of Exercise and Housing Condition on Murine Natural Killer Cell

Accepted: July 20, 1991

Acute exercise and exercise conditioning

frequency of ASGMI + splenocytes between trained and untrained animals, irrespective of presence or absence of lung tumor colonies. There was a significant effect of housing (p < 0.02), but not of physical activity, in mice with

have been shown to affect the activity of natural killer (NK)

successful tumor takes with greater numbers of group

cells as well as the growth of experimentally induced tumors in animals. Since psychosocial factors are also

housed animals (29/59) with tumor relative to individually housed animals (13/60). The data reveal that environmental factors such as housing conditions can modify the effects of exercise on natural tumor immunity. Also, the effect of exercise conditioning on NK cells in tumor and nontumor bearing animals differs significantly and, further, raises the underlying question of the physiological significance of exercise-induced changes in immune function.

known to alter NK activity and tumor growth, isolation, a known psychosocial stressor of mice, was also investigated

to see if housing condition could alter exercise-induced changes in NK cell activity and tumor growth. NK cell activity and concentration of asialo GM, (ASGM,) positive splenocytes were measured in male C3H mice inoculated i. v. with CIRAS 3 tumor cells. Mice were housed individually or in groups of four and trained to run for eight weeks

on a rodent treadmill; controls remained sedentary throughout the experimental period. At four weeks into the training protocol, mice were injected with the tumor cells and continued to run for four weeks after tumor exposure.

Introduction

In recent years there have been several reports on exercise, physical conditioning and the immune response (13, 16). For the most part, these studies have tried to link exercise-induced changes in immunity in healthy subjects with either disease risk or disease outcome. Natural killer (NK) cell activity is one such example. NK cells are a group of non-T, non-B lymphocytes, which have the unique capacity to lyse a variety of virus-infected cells as well as tumor cells without prior sensitization and which may play an important role in the control of tumor metastasis (10, 21). NK cells are thought to represent the major first line of defense against neoplasia (15). An increase in NK cell activity in highly trained subjects, at rest, was postulated to confer better resistance to infectious disease (18); indeed, a moderate exercise training program in mildly obese, premenopausal women produced an increase in Int.J.SportsMed. 13(1992)167—171 GeorgThieme Verlag Stuttgart New York

Key words

Training, natural killer cells, experimental tumors, psychosocial stress

NK cell activity which was correlated with reduced upper respiratory tract infection symptomatology (17).

Exercise-induced changes in NK cell activity have also been hypothesized to reduce cancer risk (16, 20). Data from epidemiological and experimental animal studies show associations between site specific cancer risk and exer-

cise conditioning (1, 8, 26) and between experimental tumorigenesis and exercise treatment (2, 6, 24). It is therefore reasonable to assume that if exercise-associated changes in cancer risk or cancer progression have an immunological component, then NK cells may be involved. However, experimental evidence for a linkage between exercise and cancer mediated by changes in one or more components of the immune sys-

tem is virtually nonexistent. Indeed, two recent reviews on physical activity and cancer were unable to draw definitive conclusions regarding exercise, immunity and cancer and urged future studies in this area (3, 27). Thus, while the observation that exercise and training alters immune function may

Downloaded by: NYU. Copyrighted material.

Abstract

L. Hoffman-Goetz, B. MacNeil, Y Arumugam, J. Randall Simpson

be of statistical significance, the question remains as to the biological significance of this phenomenon.

To investigate the effect of moderate exercise training on immune function and cancer, mice were trained to run on a treadmill for 8 weeks, exposed to a syngeneic tumor at the midpoint of the exercise treatment, and NK activity and tumor formation assessed after 4 weeks. In addition, we asked whether housing condition, which is well documented to aftèct tumor growth (28) and a variety of immune parameters (25) in mice, modifies exercise-associated changes in NK activity and tumor growth. This latter question has both conceptual and practical importance: in order to establish causal and mechanistic relationships among exercise, immune responses and disease, animal models are necessary. Subtle differences in background environmental 'stressors' may have effects on wmor growth and immune function which influence the interpretation of the role of the exercise treatment. Methods

Subjects Male C3H/HeNCrIBR mice (Charles River

Assay for Experimental Pulmonary Metastasis Tumor cells were harvested from subconfluent

cultures in the exponential growth phase by overlaying the cells with a thin layer of 0.5% trypsin-0.2% EDTA for 10 mm

to ensure detachment. The flask was tapped sharply to dislodge the cells, and tissue culture medium was immediately added. The cells were then washed, counted on a hemocytometer using Turk's solution as a staining agent, and resuspended in Hanks' Balanced Salt Solution at a concentration of 3 x i0 cells/mi. Tumor cell viability was about 95%, based on the ability of tumor cells to exclude trypan blue. All cell counts were performed as above except where indicated.

After 4 weeks of exercise or no exercise, all mice were injected with 200 d of CIRAS 3 (6 x l0 cells). Mice either continued to run or remained sedentary for four weeks at which point they were sacrificed for the experimental tumor metastasis assay. Intratracheal perfusions with Bouin's solution were done in order to facilitate visualizaton of lung

tumor colonies. The number of lung surface tumor colonies was determined with the aid of a dissecting microscope.

Canada, St. Constant, Quebec), initially weighing 19.9—24.6 g,

were maintained on a 12 h light/l2 h dark cycle at 21 I °C. The mice were housed in standard shoebox cages in groups of four upon arrival; the mice were group housed for a two-week acclimation period. Food (Ralston Purina rodent chow) and tap water were available ad libitum throughout the study.

Treatment Conditions Mice were randomly assigned to group (G) housing (n = 4/cage) or to individual (I) housing conditions; within each housing condition, the mice were further randomized into treadmill exercised (EX) or sedentary controls (SED). The four treatment conditions were: GxEX (n = 30), GxSED (n = 29), IxEX (n = 30) and IxSED (n = 30). The exercise protocol, similar to that previously established in our laboratory (12), consisted of run training on a motorized tread-

mill (Collins Rodent Treadmill, Braintree, Ma.) for eight weeks. An initial accomodation period of 2 weeks (build-up from 12 to 30 rn/mm, 12 mm to 30 mm, 00 slope, 5 x per week) was followed by 6 weeks of training at 30 rn/mm, 30 min/ses-

sion, final slope of 8°, 5 x per week. We have previously shown that exercise at this intensity in the C3H strain increases succinate dehydrogenase activity in skeletal muscle indicative of training (20). Mice were sacrificed 72 hours after cessation of the final training run.

Tumor

The CIRAS 3 tumor line used in the present study is an H-ras-10T1/ 2 transformed murine fibroblast line, which was a gift from Dr. A. Greenberg (Manitoba Institute of Cell Biology). This tumor cell line is known to be sensitive to

lysis by natural killer cells (11). Aliquots of tumor cells (I x 106 cells) were kept frozen at — 80 °C in DMSO until use.

The growing tumor cell line was maintained in tissue culture medium (Dulbecco's MEM-F12 medium) supplemented with 10% fetal bovine serum and 50 U Penicillin/mL. Cell cultures were maintained at 37 °C in a humidified atmosphere containing 5% C02.

Splenic Effector Cell Suspensions Spleens excised from mice were made into single-cell suspensions by pressing through fine gauze mesh with a syringe. The preparation was suspended in RPMI-l640 medium supplemented with 2 mM L-glutarnine, 10% heat inactivated fetal bovine serum, 50 U/mI each of penicillin and

streptomycin, 2.5 x 10 5M 2-mercaptoethanol, and 10 mM HEPES buffer. Erythrocytes were lysed with distilled water, the remaining cell suspension washed with 2 x PBS, and re-

suspended in RPMI-1640 at a concentration of 1.5 x l0 cells/mI for the microcytotoxicity assay and at 2.0 x l0 for the immunofluorescence staining. Spleen suspensions were not available for four animals.

Target Cells The YAC-l cell line, a Moloney leukemia virus

induced lyniphorna of the A/Sn strain of mouse, was maintained in suspension in RPMI-1640 medium supplemented with penicillin, streptomycin, glutamine and 10% heat inactivated fetal bovine serum. YAC-I were used as highly sensitive targets for NK mediated cytolysis in vitro.

Cytotoxicity Assay A standard microcytotoxicity assay was used as described (18). This assay tests the ability of NK cells from the effector cell population to lyse radiolabelled tumor cell targets releasing the radiolabel into the supernatant. The radioactivity of the supernatant is then measured to determine the extent of target cell lysis. Briefly, target cells were labelled with

sodium chromate [5tCr], washed, and distributed into Vbottom 96 well microtiter plates at a concentration of I 0' cells/i 00 Ill. Effector spleen cells at various effector-to-target (E : T) ratios ranging from 150: ito 18: 1 were added in 100 il volumes in triplicate to the plated labelled tumor target cells. For the spontaneous release control, 100 il of medium was added to the labelled target cells. For the total 51Cr release 1

control, 100 il of Triton-X (detergent) was added to the

Downloaded by: NYU. Copyrighted material.

168 mt. J. Sports Med. 13 (1992)

mt. J. Sports Med. 13(1992) 169

Exercise and Ho using Condition on Murine NaturalKiller Cd/Activity and Tumor Growth A

Table 1 NK cytotoxicity levels in mice four weeks after intravenous injection of 6 x io CIRAS 3 tumor cells expressed as a percent of maximal 51Cr release at the effector to target ratio of 150:1 and displayed by main effect treatment conditions and by the presence or absence of visible tumors % Lysis at E:T Ratio of 150:1

Tumor

a

U G x EX o G x SED

• I x EX o IxSED

0

Present

No Tumor Present

0>'

16.8± 1.7

17.7±1.2*

z

EX

17.4± 1.0

SED

13.9±1.1

14.7±2.2

13.4±1.2

G

15.7± 1.1

16.3± 1.8

15.0±1.3

I

15.6± 1.0

14.4±2.1

15.9± 1.2

Activity

20

0

Condition Combined Groups

visible Tumors

10

5

I-lousing

amean±standard error of the mean. *significantly different from SED group; p < 0.02.

150:1

37:1

18:1

E:T Ratio

B

in a 5% C02 environment for 4—6 h. After incubation, the plates were centrifuged at 400 g for 10 mm, and 100 .tl of the supernatant removed from each well and counted on a Beckman 5500 gamma counter. Data were expressed as percent specific 51Cr release:

no visible Tumors

20

• G x EX El

>'

GxSED

• I x EX

0 15 0 >'

o IxSED

0 x 100%

Z

10

cpm total release — cpm spontaneous release

Invnunofluorescence The frequency of splenocytes positive for the glycosphingolipid asialo GMi (ASGMi) was determined by

0

indirect immunofluorescence. Fifty microliters of cell suspension (2 x l0 cells) were incubated with 5 1iJ of rabbit ASGMi

(Cedarlane Laboratories, Hornby, Ontario) in 45 l of medium for 45 mm at 4 °C. The cells were washed with medium

and incubated with 100 of a 10 .tg/ml solution of goat antirabbit IgG complexed to fluorescein isothiocyanate for an additional 45 mm in the cold. Stained cell suspensions were fixed with 2 g% paraformaldehyde and scored for immunofluorescence by microscopy.

150:1

75:1

37:1

18:1

E:T Ratio Fig. 1 NK cytotoxicity levels in mice four weeks after intravenous injection of 6 x io CIRAS 3 tumor cells expressed as a percent of maximal 51Cr release and displayed by treatment condition at various effector to target ratios. Panel A shows NK responses in mice with visible lung tumor colonies present; panel B shows NK responses in mice with no visible lung tumorcolonies present (*significantly different from SED groups with no visible tumors; p

Differential effects of exercise and housing condition on murine natural killer cell activity and tumor growth.

Acute exercise and exercise conditioning have been shown to affect the activity of natural killer (NK) cells as well as the growth of experimentally i...
118KB Sizes 0 Downloads 0 Views