The pituitary gland is required for protection against lethal ... - PNAS

Proc. Nati. Acad. Sci. USA Vol. 88, pp. 2274-2277, March 1991

Immunology

The pituitary gland is required for protection against lethal effect's of Salmonella typhimurium (growth hormone/interfero'n y/macrophage activation/free radicals/immunophysiology) CARL K. EDWARDS III*t, LIBBY M. YUNGER*, ROBERT M. LORENCE*t, ROBERT DANTZER§, AND KEITH W. KELLEY*¶

*Laboratory of Immunophysiology, Department of Animal Sciences, University of Illinois, 162 ASL, 1207 West Gregory Drive, Urbana, IL 61801; and 1lnstitut National de la Recherche Agronomique-Institut National de la Sant6 et de la Recherche Mddicale, Unitd de Recherches de Neurobiologie des Comportements, U 176, Rue Camille-Sain't-Saens, 33077 Bordeaux, France Communicated by Marvin P. Bryant, December 20, 1990 (received for review November 8, 1990)

intermediates that can kill pathogenic microbes (16). We now report that hypophysectomized rats are much more susceptible than sham-operated rats to the lethal effects of virulent S. typhimurium and that GH is as effective as IFN-y in increasing their survival. The beneficial role of GH and IFN-y is mediated through activation of macrophages to an antibacterial state, an effect that is blocked by scavengers of toxic oxygen intermediates.

One-half of pituitary-intact or sham-operated ABSTRACT rats survive infection with 10 colony-forming units of Salmonella lyphimurium, whereas rats without a pituitary gland all die within a few days. When the dose of S. typhuimwwim is reduced 600-fold, 15-25% of the hypophysectomized rats survive, and the survival rate is sgicatyenhanced by administration of tetracycline, recombinant interferon y (IFN,y), or recombinant growth hormone (GH). The protective effect of Gil is abolished by heat inactivation or with an antibody to Gil. Spleens from normal and hypophysectomized rats treated with tetracycline, IUFN-y, or Gil have 59-99% fewer bacteria 5 dlays after infection as compared to control rats. Peritoneal macrophages from hypophysectomized rats that are infected in vitro with S. atyphimurium kill half as many extracellular bacteria as compared to pituitary-intact rats, and this bactericidal capacity is sgicatly augmented 75-95 % by either Gil or IUFN-,y. These data establish that the pituitary gland 'i's essential for hoineostasis during an infectious -episode and that Gil plays an important role in host resistance by augmenting the ability of macrophages to kill S. typhimurium.

MATERIALS AND METHODS Rats. Six-week-old female albino Sprague-Dawley rats (Johnson Laboratories, Bridgeview, IL) were hypophysectomized or sham-hypophysectomized at least- 15 days prior to initiation of experimental treatments. Completeness of hypophysectomy was determined as described (14, 17, 18). Host Protection Experiments. Rats were challenged intraperiton'eally (i.p.) with virulent S. typhimurium bacteria (strain 84-4728, kindly supplied by B. 0. Blackburn, U.S. Department of Agriculture Infectious Disease Center, Ames, IA) grown in trypticase soy agar broth (BBL Microbiology Systems). Pituitary-intact rats were injected i.p. with S. ty-phimurium on day 0 following subcutaneous (s.c.) treatment for 6 days with either 200 1LI of buffer (placebo; 0.15 M NaCl/0.03 M NaHCO3, pH 9.5), 1000 units of recombinant rat IFN-y (specific activity, 4 x 106 units/mg of protein; Amgen 'Biologicals), 500 pAg of normal or heat-inactivated pituitaryderived porcine G'H (A. F. Parlow, University of CaliforniaLos Angeles Medical Center, Torrence), or 100 mg of tetracycline per kg of body weight (Sigma). These treatments continued for another 6 days after infection. Experiments with hypophysectomized rats were conducted in a similar manner, except the daily dose of IFN-y, porcine GH, or heat-inactivated porcine GH was reduced to 500 units, 48 .Ag, or 48 Ag, respectively. Recombinant porcine GH (PitmanMoore, Terre Haute, IN) wa's also tested in hypophyse'ctomized rats according to the same protocol. All reagents were determined to be free of endotoxin (assay sensitivity of 25 pg/mi), as described (14, 15, 17). Mortality was recorded at 12-hr intervals. Animals that were moribund for >24 hr were euthanized, in accordance with guidelines established by the University of Illinois Campus Laboratory Animal Care Advisory Committee.

A central tenet of neuroendocrine-immune system interactions is that bidirectional communication signals exist that 'are important in host resistance to infectious, autoimmune, and neoplastic diseases (1). In support of this hypothesis, it has recently been demonstrated that a genetic defect leading to aberrations in signaling- within the immune-central nervous system-pituitary axis causes an increase in susceptibility to autoimmunie diseases in animal models of arthritis (2), multiple sclerosis (3), and Hashimoto thyroiditis (4). Suppression of the synthesis of pituitary prolactin increases the susceptibility of mice to Listeria monocytogenes and inhibits tumoricidal activity of macrophages (5). However, there are no experimental data that demonstrate that the pituitary gland controls the susceptibility of animals to infectious diseases by altering the ability of cells of the immune system to kill

microorganisms. Salmonella species are an important cause of gastroenteritis in man and animals and are often used to study typhoid immunity in rodents. These pathogens are capable of surviving within phagocytic cells (6) and macrophages are needed for host defense against this microorganism (7). Macrophages must be exposed to cytokines, such as interferon 'y (IFN-y) (8)1, granulocyte/macrophage colony-stimulating factor (9), or tumor necrosis factor a' (10), to fully exert their antibacterial activity (11) and protect animals against lethal infection with Salmonella typhimurium. All of these substances (12, 13), as well as growth hormone (GH; refs. 14 and 15), prime macrophages and neutrophils to release reactive oxygen

Abbreviations: GH, growth hormone; IFN-y, interferon y; SOD, superoxide dismutase; cfu, colony-formuing unit(s). tPresent address: Department of Immunology, Marion Merrell Dow Inc., Marion Park Drive, P.O. Box %627, Kansas City, MO 641340627. tPresent address: Department of Pathology, Rush-Presbyterian-St. Luke's Medical Center, 1753 West Congress Parkway, Chicago, IL

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

60612.

I~whom reprint requests should be addressed. 2274

Immunology: Edwards et A

Proc. Natl. Acad. Sci. USA 88 (1991)

Neutralization of GH Activity. A specific, heat-inactivated (30 min at 560C) guinea pig antiserum to porcine GH that was kindly provided by Dennis Marple (19) as well as heatinactivated normal guinea pig serum (Pel-Freeze Biologicals) were incubated with native, pituitary-derived porcine GH for 1 hr at- 370C before injection into rats. Native, pituitaryderived porcine GH was heat-inactivated by incubation for 1 hr at 70'C. Measurement of S. typhimurum in Vivo. Five days after infection with S. typhimurium in vivo, spleens were removed aseptically and individually homogenized in 0.9%o NaCl (pH 7.4), and the number of bacterial colony-forming units (cfu) in the homogenates was determined. Bacterial Killing by Macrophages. Bactericidal activity of peritoneal macrophages, harvested from normal and hypophysectomized rats that had not been previously exposed to S. typhimurium in vivo but had been treated for 6 days at the indicated concentrations with either buffer (placebo), IFN-y, or pituitary-derived porcine GH, was determined by inhibition of growth of S. typhimurium. Peritoneal macrophages from treated rats were collected and isolated by adherence to plastic as described (14, 20). Macrophages (7 x 105 cells) were then incubated for 15 hr in Labtek chambers in antibiotic-free RPMI 1640 medium (supplemented with 1% low endotoxin fetal bovine serum; HyClone) in a volume of 750 1.l at 37C, 5% CO2 in air atmosphere. Macrophage cultures were then washed with RPMI 1640 medium at 370C. Eight hundred microliters of antibiotic-free RPMI 1640 medium/5% fetal bovine serum was added to each culture along with 100 1l of superoxide dismutase (SOD, 2.5 mg/ml; ICN), catalase (2.5 mg/ml; Sigma), or medium. Three hours later, 100 p1l of S. typhimurium (5 x 106 cfu) was added that had been opsonized with normal rat serum for 20 min at 37C. Bacteria were opsonized because we have previously shown that GH-treated phagocytic cells require a triggering stimulus, such as opsonized-zymosan, formyl-methionylleucylphenylalanine, or phorbol 12-myristate 13-acetate, to augntact

2275

ment the secretion of superoxide anion into the medium (14, 15, 20). After 8 hr, total cfu of S. typhimurium in the supernatant was determined by serial dilutions on trypticase soy agar. Statistics. Host protection data were analyzed using a contingency x2 analysis test (21). A general linear models procedure was performed on the data using SAS, and differences between treatments were assessed by Duncan's new multiple range test (21).

RESULTS Hypophysectomy and Resistance to S. typhimurum. The amount of S. typhimurium was established from preliminary experiments that showed that hypophysectomized rats were much more susceptible to doses of S. typhimurium that only mildly affected pituitary-intact rats. Injection of 1200 X 106 cfu i.p. caused 47% of normal and sham-operated rats to die within 7 days, whereas this dose of bacteria rapidly killed all of the hypophysectomized rats (P < 0.01, x2 test; Fig. 1). The number of cfu had to be reduced to 2 x 106 to permit -survival of 27% of hypophysectomized rats. However, this dose of bacteria (2 x 106) led to 100%o and 87% survival of normal and sham-operated rats, respectively, and both of these values were significantly greater (P < 0.001, x2 test) than that of hypophysectomized rats. In the absence of an infection with S. typhimurium, all rats survived. Host Resistance in Rats Treated with GH, FN-y, or Tetracycline. Pituitary-derived porcine GH, recombinant porcine GH, recombinant rat IFN-y, or tetracycline was administered for 6 days before and after infection with S. typhimurium to pituitary-intact (Fig. 2A) or hypophysectomized (Fig. 2B) rats. Forty-four percent (number of survivors/number tested; 16/36) of pituitary-intact rats were alive 7 days after infection with S. typhimurium (360 x 106 cfu) and treated with control buffer (placebo), whereas a dose of 500 ,ug of pituitary-derived GH per day, 1000 units of recombinant rat

(100)

100

C')

550 cJa)

a)

0

O X 2 X 10

F

~~~~~ypox

W(0~~~~~~~~~~~~~~~~~~~~~~~~()

6CFUDose of Salmonella typhimurium1200 x 106CFU

FiG. 1. Removal ofthe pituitary gland increases susceptibility to the lethal effects of S. typhimurium. Fifteen rats were used in each treatment group, for a total of 90 rats. The value of parentheses indicates the actual percentage survival 7 days after challenge with S. typhimurium. The asterisks indicate that survival of hypophysectomized (Hypox) rats was significantly less (P < 0.01) than the other two groups, as assessed by X2 analysis.

2276

Immunology: Edwards et al.


A

B loo

v

2

cit

CO

CD c

U 0 4)

CD 0C

I

Placebo

rrlFN-y **--- GH -o- H-l GH --

Tetracycline 2

4

6

8

10

12

14

Days After Injecting Salmonella typhimurium

Days After Injecting Salmonella typhimurium

FIG. 2. GH and IFN-y increase survival of pituitary-int. (A) and hypophysectomized (B) rats following challenge with S. typhimurium on day 0. The number of rats that were used, 7-day survival -ates, and statistical analysis by x2 test at day 7 are presented in the text. rrIFN-y, recombinant rat IFN-y; GH, native, pituitary-derived porcine GH; HI-I GH, heat-inactivated native, pituitary-derived porcine GH.

IFN-'y per day, or 100 mg of tetracycline per kg significantly increased (P < 0.05) the survival rate to 75% (18/24), 83% (20/24), or 79o (19/24), respectively. The enhancing effect of GH (500 pg/day) on survival rate of normal rats infected with S. typhimurium was abrogated by heat'inactivation [44% survival in the placebo group versus 54% (13/24) in the treated group], a treatment that reduces the growth-restoring properties of GH in hypophysectomized rats by =60%o (17). Although only 13% (2/15) of the hypophysectomized rats infected with S. typhimurium (0.5 x 106 cfu) survived after 7 days, administration of 500 units of recombinant rat IFN-y per day, 48 Jug of pituitary-derived porcine GH per day, or 100 mg of tetracycline per kg significantly increased the survival rate (P < 0.05, Fig. 2B) to 60o (9/15), 53% (8/15), and 73% (11/15), respectively. Heat-inactivation of GH administered at a dose of 48 pug/day abrogated (P < 0.05) the increase in survival [13% survival in the placebo group versus 20o (3/15) in the heat-inactivated group]. Coadministration of a guinea pig antiserum specific for porcine GH (19) reduced (P < 0.05) the protective effects of this molecule (4/15 for 27% survival) as compared to treatment with porcine GH in combination with a control guinea pig serum (10/15 for 67% survival). The specific guinea pig antiserum alone did not affect survival of hypophysectomized rats (3/15 for 20o survival). Furthermore, the beneficial effect of GH was not mimicked by injecting an equivalent amount (48 pug daily) of an irrelevant protein, porcine serum albumin (1/12 for 8% survival). To verify that the effect of pituitary-derived porcine GH was not due to contaminating proteins, graded doses of recombinant porcine GH were administered to a completely different group of hypophysectomized rats according to the same protocol. Seven days after infection with 7 x 106 cfu, survival rates of rats injected with 24, 48, and 96 ug of recombinant porcine GH per rat per day were 33% (4/12), 67% (10/15), and 67% (8/12), respectively. All doses of recombinant porcine GH, as well as the pituitary-derived preparation at 48 pg/day (8/15 for 53% survival) and recombinant rat IFN-y at 500 units/day (8/15 for 53% survival),

significantly increased (P < 0.01) survival rates compared to placebo-treated rats (0/15 for 0%6 survival). S. typhimuium in Spleens of Rats Treated with GH, IFN-y, or Tetracycline. The number of cfu in the spleens of surviving rats was assessed 5 days after infection. In pituitary-intact and hypophysectomized rats, tetracycline reduced (n = 3; P < 0.05) splenic cfu burden by 99%6. Similarly, IFN-y and pituitary-derived porcine GH significantly reduced cfu in the spleens of intact (79%o and 59%o, respectively; P < 0.01) and hypophysectomized (95% and 89%, respectively; P < 0.001) rats. Heat-inactivation of GH partially attenuated the reduction in splenic bacterial burden in intact (19o) and hypophysectomized (58%, P < 0.05) rats. Killing of S.. typhimurium by Macrophages in Vitro. In the absence of SOD or catalase, all macrophage preparations reduced (P < 0.05) the number of bacteria recovered in the extracellular medium (Table 1) compared to cultures without macrophages (9.0 ± 1.5 x 106 cfu). Furthermore, recovery of extracellular bacteria in cultures of macrophages derived Table 1. GH injections increase killing of S. typhimurium by peritoneal macrophages from pituitary-intact and hypophysectomized (Hypox) rats infected in vitro cfu x 10-6 In vivo treatment Intact Hypox 3.94 + 0.16t 2.15 ± 0.28* Placebo + SOD 8.17 + 0.690§ 4.71 ± 0.54tt + catalase 7.66 ± 0.72t 6.00 ± 0.58§ 1.06 ± 0.2411 0.12 ± 0.01 IFN-y + SOD 9.10 ± 0.55*§ 2.94 ± 0.22*11 9.25 ± 0.640§ + catalase 3.81 ± 0.22tll 0.81 ± 0.3211 0.19 ± 0.041 Porcine GH + SOD 3.25 + 0.4111 5.08 + 0.60t + catalase 5.45 ± 0.46*1 9.69 ± 0.70§ n = 5 rats in each treatment. Means within either the intact or Hypox columns with different superscripts are different (P < 0.05). Additionally, cfu were lower (P < 0.05) in intact as compared to Hypox groups for the placebo, IFN-y, and porcine GH treatments.

Immunology: Edwards et al. from placebo-treated, pituitary-intact rats was 2-fold less than that from hypophysectomized rats (Table 1; P < 0.05). A similar significant reduction in cfu from macrophages of intact as compared to hypophysectomized rats was observed even when rats were treated in vivo with IFN-y or porcine GH (P < 0.05, Table 1). To our knowledge, it has not been demonstrated previously that macrophages from hypophysectomized rats are less effective in killing extracellular S. typhimurium than those derived from intact rats. Treatment in vivo with either IFN-y or GH at the doses described above significantly enhanced by ;85% the ability of macrophages isolated from either intact or hypophysectomized rats to reduce the number of extracellular bacteria in vitro (Table 1). Toxic oxygen intermediates play a role in this bactericidal activity of activated macrophages because SOD and catalase significantly blocked (P < 0.05) the response in all treatment groups.

DISCUSSION Our finding that the pituitary gland was required for rats to resist the lethal effects of S. typhimurium has important implications for understanding the complex interactions that occur between the immune and central nervous system following exposure to a pathogenic bacterium. We recently demonstrated that GH is as effective as IFN-y in priming rat, porcine, and human macrophages or neutrophils for the secretion of superoxide anion (14, 15). These data and others (reviewed in refs. 22 and 23) suggested that GH might act in vivo to protect animals against lethal bacterial infections in a manner similar to IFN-y. Data presented in this report supported this hypothesis by showing that natural and recombinant GH increased survival rates of intact and hypophysectomized rats after infection with S. typhimurium. This protective effect was abrogated by a specific antibody to GH or by heat-inactivation of the GH molecule. The finding that macrophages from either pituitary-intact or hypophysectomized rats treated in vivo with GH were activated to an antibacterial state that could be blocked by scavengers of toxic oxygen intermediates (SOD and catalase) probably explains the significant reduction in splenic bacterial burden of these rats. However, other cell types could also be involved, since GH can enhance the activity of rat natural killer cells (24, 25) and granulocytes (15), and both of these cell types have been recently suggested to be effector cells in killing S. typhimurium (26). Since GH and the closely related molecule prolactin (5) increase survival of animals infected with two different pathogenic bacteria (S. typhimurium and L. monocytogenes), members of the somatomammotrophic hormone gene family are critical for host resistance to bacterial infections. Recombinant human GH is finding increased use in human medicine, ranging from augmenting growth rate in GHdeficient children to increasing muscle mass in aged men (27). GH may also be used in the animal food chain for humans (28). However, even though GH augments a number of immune events in young and old animals (14, 15, 17, 18, 20, 22-25), it was not known whether GH could also affect the resistance of animals to infectious disease. The present data offer strong support for the idea that GH secretion by the neuroendocrine system acts in concert with phagocytic cells


2277

of the immune system to maintain host resistance to infectious diseases. We thank D. Segre, S. Arkins, R. Franklin, H. Lewin, L. Schook, and D. Kranz for helpful suggestions with this manuscript. This research was supported in part by National Institutes ofHealth Grant AG06246, Office of Naval Research Grant N00014-89-J-1956, U.S. Department of Agriculture Grant 89-37265-4536, Pitman-Moore, Inc., Moorman Manufacturing Company, and Illinois Agriculture Experiment Station Project 20-0386 (to K.W.K.).

1. Blalock, J. E. (1989) Physiol. Rev. 69, 1-32. 2. Sternberg, E. M., Hill, J. M., Chrousos, G. P., Kamilaris, T., Listwak, S. J., Gold, P. W. & Wilder, R. (1989) Proc. Natl.

Acad. Sci. USA 86, 2374-2378. 3. Mason, D. W., MacPhee, I. A. M. & Antoni, F. A. (1990) Immunology 70, 1-5. 4. Schauenstein, K., Fassler, R., Dietrich, H., Schwarz, S., Kromer, G. & Wick, G. (1987) J. Immunol. 139, 1830-1833. 5. Bernton, E. W., Meltzer, M. S. & Holaday, J. W. (1988) Science 239, 401-404. 6. Buchmeier, N. A. & Heffron, F. (1990) Science 248, 730-732.

7. Mackness, G. B. (1971) J. Infect. Dis. 123, 439-445. 8. Matsumura, H., Onozuka, K., Terada, Y., Nakano, Y. & Nakano, M. (1990) Int. J. Immunopharmacol. 12, 49-56. 9. Morrissey, P. J. & Charrier, K. (1990) J. Immunol. 144, 557561. 10. Nakano, Y., Onozuka, K., Terada, Y., Shinomiya, H. & Nakano, M. (1990) J. Immunol. 144, 1935-1941. 11. Adams, D. 0. & Hamilton, T. A. (1987) Immunol. Rev. 97, 5-27. 12. Reed, S. G., Nathan, C. F., Pihl, D. L., Rodricks, P., Shanebeck, K., Conlon, P. J. & Grabstein, K. H. (1987) J. Exp. Med. 166, 1734-1746. 13. Nathan, C. F., Prendergast, T. J. & Wiebe, M. E. (1984) J. Exp. Med. 160, 600-605. 14. Edwards, C. K., III, Ghiasuddin, S. M., Schepper, J. M., Yunger, L. M. & Kelley, K. W. (1988) Science 239, 769-771. 15. Fu, Y.-K., Arkins, S., Wang, B.-S. & Kelley, K. W. (1991) J. Immunol. 146, 1602-1608. 16. Babior, B. M. (1978) N. Engl. J. Med. 298, 659-668. 17. Edwards, C. K., III, Lorence, R. M., Dunham, D. M., Arkins, S., Yunger, L. M., Greager, J. A., Walter, R. J., Dantzer, R. & Kelley, K. W. (1991) Endocrinology 128, 989-9%. 18. Kelley, K. W., Brief, S., Westly, H. J., Novakofski, J., Bechtel, P. J., Simon, J. & Walker, E. B. (1986) Proc. Natl. Acad. Sci. USA 83, 5663-5667. 19. Marple, D. N. & Aberle, E. D. (1972) J. Anim. Sci. 34, 261265. 20. Davila, D. R., Edwards, C. K., III, Arkins, S., Simon, J. & Kelley, K. W. (1990) FASEB J. 4, 2906-2911. 21. SAS Institute (1987) SAS User's Guide: Statistics (SAS Institute, Cary, NC). 22. Berczi, I. (1989) Dev. Compar. Immunol. 13, 329-341. 23. Kelley, K. W. (1989) Biochem. Pharmacol. 38, 705-713. 24. Davila, D. R., Brief, S., Simon, J., Hammer, R. E., Brinster, R. L. & Kelley, K. W. (1987) J. Neurosci. Res. 18, 108-116. 25. Saxena, Q. B., Saxena, R. K. & Adler, W. H. (1982) Int. Arch. Allergy Appl. Immunol. 67, 169-174. 26. Langermans, J. A. M., van der Hulst, M., Nibbering, P. H. & van Furth, R. (1990) J. Immunol. 144, 4340-4346. 27. Rudman, D., Feller, A. G., Nagraj, H. S., Gergans, G. A., Lalitha, P. Y., Goldberg, A. F., Schlenker, R. A., Cohn, L., Rudman, I. W. & Mattson, D. E. (1990) N. Engl. J. Med. 323, 1-6. 28. Juskevich, J. C. & Guyer, C. G. (1990) Science 249, 875-884.