- Email: [email protected]
Immunology discovers physiology
'~" ~" ,~
Veterinary immunology .
E LS E VIE R
and
Veterinary Immunology and Immunopathology 43 (1994) 157-165
immunopathology
Immunology discovers physiology Keith W. Kelley a'*, Rodney W. J o h n s o n b, R o b e r t D a n t z e r c aLaboratorv of lmmunophysiology, 207 Plant and Animal Bioteehnology Laboratory, Department of Animal Sciences, University oflllinois, 1201 West Gregory Drive, Urbana, IL 61801, USA b323 Animal Sciences Laboratory, Department of Animal Sciences, University oflllinois, 1201 West Gregory Drive, Urbana, IL 61801, USA CLaboratory of lntegrative Neurobiology, INRA-INSERM, Rue Camille Saint-Sa£;ns 33077 Bordeaux Cedex, France
Abstract
Not so long ago, it was believed that the brain is totally devoid of immunologic reactions, that cytokines derived from activated leukocytes serve only as communication molecules between leukocytes and that the immune system is regulated solely by intrinsic mechanisms. One by one, these old-time, traditional views have fallen by the wayside as neuroscientists, endocrinologists and pharmacologists have begun to explore immunology. The old view was that the immune system is autonomous because it neither affects nor is it affected by other physiologic systems. The new view is that cells of the immune system are inextricably linked with other physiological systems, including the neuroendocrine, cardiovascular, reproductive and central nervous systems (CNS). Changes in one system evoke changes in the other, and it is likely that communication loops have evolved between cells of the immune system and those of other tissues to coordinate and regulate functional activities aimed at preserving homeostasis during inflammation. The integrated view of immunophysiologists that cells of the immune system interact with the entire body, rather than existing as a separate physiologic system that operates autonomously, should help to unravel a number of mysteries in immunoregulation, such as the well-recognized redundant and pleiotropic properties of cytokines. Unfortunately, very few of these ideas have been incorporated into studying immunity of domestic animals. A complete understanding of immunobiology will be achieved only after this new field of immunophysiology is integrated into current immunological thinking. The purpose of this short article is to describe new discoveries which provide insights into how leukocytes discriminate between self and non-self by enlisting the aid of the neuroendocrine and CNS and to document what is known about the immunophysiology of pigs. *Corresponding author. Tel. (217) 333-5141; fax (217) 244-5617. 0165-2427/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0165-2427 ( 94 ) 06020-Z
158
K. V~ Kelley et al. / Veterinary Immunology and Immunopathology 43 (1994) 15 7-165
1. Introduction: the new view of the brain as an immunoprivileged site
The CNS has been considered to be immunologically privileged because it: (a) lacks lymphatic drainage; (b) has relatively impermeable, non-fenestrated cerebral capillary endothelium forming tight junctions (no gap between endothelial cells). The blood-brain barrier is actually a combination of several physical barriers, including the close juxtaposition of endothelial cells, the abundance of astrocytes in the surrounding cerebral capillaries and the small size of the extracellular space; (c) displays very little expression of either class I and II major histocompatibility complex (MHC) molecules; (d) readily accepts allografts. However, this traditional view of the brain as an immunoprivileged site is changing. First, antigens injected into the brain can be found in deep cervical lymph (Yamada et al., 1991 ), apparently escaping the CNS via the cranial nerves, which indicate that there may be substantial traffic from the CNS to lymph. Second, T lymphoblasts do cross the blood-brain barrier, probably by secreting specific endoglycosidases, and do so regardless of MHC phenotype or type of activation signal (Hickey et al., 1991 ). Third, large molecules such as cytokines communicate with the CNS, as demonstrated by i.p. injections of IL-1 causing an elevation in hypothalamic set-point leading to fever. Cytokines could gain access to the CNS via several mechanisms: (a) through circumventricular organs of the brain (which lack a true blood-brain barrier); (b) using glial cells to send messages via prostaglandin E2 to various parts of the CNS; (c) utilizing specific transporters in the cerebral vascular endothelium; or (d) acting indirectly by inducing their own synthesis within the CNS (discussed by Kent et al., 1992a). Finally, perivascular microglia, which develop from the bone marrow and are considered to be a subtype of macrophages in the CNS, can be readily induced to express class II MHC and function as important antigen-presenting cells in experimental allergic encephalomyelitis (Hickey and Kimura, 1988 ). Interferon-y (IFN-7) induces expression of class I MHC molecules on microglia, astrocytes and oligodendrocytes, and class II proteins on microglia and astrocytes (reviewed by Benveniste, 1992). Encephalitic viruses such as lymphocytic choriomeningitis virus persist because neurons lack sufficient expression of class I MHC molecules to effectively engage virus-specific cytotoxic T lymphocytes (Joly et al., 1991 ). These findings are consistent with the new view that immune privilege is an active rather than a passive process that is controlled, in part, by specific suppression of inflammatory responses in the CNS (Cserr and Knopf, 1992). Transforming growth factors ill, if2 and f13 are expected to play important roles in this regard because they are potent suppressive cytokines and they are secreted in both the latent and active forms by microglia and astrocytes (Constam et al., 1992; Cunha et al., 1993). Synthesis of TNF-o~ by lipopolysaccharide (LPS)stimulated porcine alveolar macrophages is significantly suppressed by nanomolar concentrations of TGF-fl2 (Dunham et al., 1990). Expression of class II MHC proteins on IFN-7 treated human microglia is potently down-regulated by colony
K. W. Kelley et al. / Veterinary lmmunology and lmmunopathology 43 (1994) 157-165
159
stimulating factor-I (CSF-1; Lee et al., 1993), which provides another means of suppressing immunologic reactions in the CNS. Leukemia inhibitory factor (LIF)/cholinergic differentiation factor (CDF) is also an important CNS-modulating cytokine (reviewed by Patterson and Nawa, 1993 ). 2. Cytokines and their receptors in the central nervous system Human patients injected with recombinant cytokines often experience chills and have neurological symptoms such as memory loss, drowsiness, weakness and motivational defects (reviewed by Dantzer and Kelley, 1989; Kelley and Dantzer, 1990; Fig. 1 ). Fever, anorexia, sleepiness and lethargy also occur in food animals injected peripherally with recombinant cytokines (Blecha, 1991; F. Blecha, perCENTRAL EFFECTS L" [
Activate Pituitary-Adrenal Axis Increase Thermogenesis Increase Body Temperature Increase Slow-Wave Sleep Sickness Behavior Decrease Body Care Aclwities | Malaise ,~ Decrease Locomotor Activlly Decrease Social Exploration Decrease Food Intake XDecrease Food Motivated BehavLor
._1 < rr I-Z LU
?
4,
LIVER
_1 < rr UJ T
Release of Acute Phase Proteins
JOINTS Pain
13._
cr LU 12.
Immune
Activators
T
GASTROINTESTINAL TRACT Inhibit Gastric Emptying Inhibit Gastric Acid Secretion
t
Hormonal I Effects
.
Cytokines ,~~._ ~,C:~,,w ot ~
I
Autocrine and Paracrine Effects
Immune Ceils
Fig. 1. Proinflammatory cytokines act centrally in the brain. Cytokines amplify not only the immune response in regional lymph nodes but also inform the central nervous system the immune system has been activated. The brain is involved in coordinating a variety of systemic responses, such as inactivity, anorexia, somnia and fever, that promote homeostasis following infection (from Kent et al., 1992a).
160
K. W, Kelley et al. / Veterinary lmmunology and lmmunopathology 43 (1994) 157-165
sonal communication, 1993 ). Since proinflammatory cytokines such as IL-1 and TNF can cause all of these symptoms, we and others have proposed that cytokines act as hormones to inform the CNS that a foreign agent has gained access to the body (Blalock, 1984; Kent et al., 1992a). Although the mechanisms by which peripheral cytokines cause animals to experience clinical signs of sickness are just now beginning to be investigated, we recently used a specific antagonist of the type I IL-1 receptor to show that induction of fever does not share the same receptor mechanisms as those responsible for the decrease in food-motivation and in social activities which are characteristic of sickness (Kent et al., 1992b). Endotoxin induces the synthesis of IL-1 by macrophages in the meninges, choroid plexus, perivascular cells and ramified microglia of the rat brain (Van Dam et al., 1992) and causes a number of behavioral abnormalities in pigs and birds (Johnson et al., 1993a,b; 1994a,b). Trauma-induced IL- I a and T N F a protein (Tchelingerian et al., 1993 ), as well as LPS-induced transcripts for IL- 1o~and ILl fl (Ban et al., 1992), are distributed in specific areas of the brain, suggesting discrete functions for these molecules in the CNS. Sympathetic neurons synthesize and secrete IL-lfl (Freidin et al., 1992). Indeed, a number of cytokines or their transcripts can be detected in the CNS constitutively or during various disease states, including IL-1, IL-2, IL-3, IL-6, CSF-1, IFN-a, IFN-fl, IFN-7, TNF-o~ and TGF-fl (reviewed by Benveniste, 1992), and these cytokines may be key players in both normal and pathogenic states, including debilitating CNS diseases such as AIDS (Morganti-Kossmann et al., 1992). It is particularly interesting that neurotrophic factors such as CDF/LIF, ciliary neurotrophic factor and oncostatin M share significant structural motifs with IL-6 and G-CSF (Bazan, 1991 ). Indeed, the striking similarities between differentiation of cells within the hematopoietic and nervous systems suggest that cytokines evolved as key regulators in both systems (Patterson and Nawa, 1993 ), and that these proteins could be particularly important for differentiation of the newly discovered neuronal stem cells (Mehler et al., 1993 ). Specific receptors for IL-1 have been identified in the hippocampus and dendate gyrus of the mouse brain (Ban et al., 1991; Cunningham et al., 1992), and we have partially cloned both the type I and II IL-1 receptors from the murine adenohypophysis and brain (Parnet et al., 1993; 1994). The pituitary gland also responds to IL-1, IL-2 and IL-6 (Woloski et al., 1985; Brown et al., 1987 ). Receptors for TNF (Kinouchi et al., 1991 ) and IL-6 (Cornfield and Sills, 1991 ; Schobitz et al., 1992 ) are found in the brain, but the specific cell types expressing these receptors are as yet unknown. It is interesting that both the brain and leukocytes express receptors for exogenous (morphine) and endogenous (fl-endorphin) opiates. It appears that porcine natural killer (NK) and phagocytic cells possess naloxone-sensitive opiate receptors because methadone suppresses free radical and NK cytolytic activity in vitro (Molitor et al., 1992a), whereas morphine impairs porcine contact sensitivity reactions in vivo (Molitor et al., 1992b). It has long been known that sick animals do not eat well, and we (Kelley et al., 1993a) have proposed that an immunological mechanism at least partially explains this phenomenon because: (a) sick animals are anorectic; (b) specific re-
K. W. Kelley et al. / Veterinary Immunology and Immunopathology 43 (1994) 15 7-165
161
ceptors for cytokines are found in the CNS; (c) several bacterial and viral pathogens induce synthesis and secretion of cytokines by myeloid cells; (d) injection of recombinant cytokines cause animals to stop eating; (e) the anorectic effect of at least one cytokine (IL-1 ) is blocked by a specific receptor antagonist. We think that cytokines are somehow involved in normal brain functions that control economically important traits of farm animals, such as food intake.
3. The pituitary gland as a link between the immune and central nervous systems
Relationships between the immune system and the brain are truly bidirectional in the sense that not only products from the immune system affect the brain, but also hormones and neurotransmitters modulate a number of immune events. This is particulary true for somatotropin (growth hormone), a hormone secreted by the anterior pituitary gland. In the absence of a pituitary gland (hypophysectomy), we have shown that rats infected with Salmonella typhimurium die much more rapidly than their controls with a hypophysis (Edwards et al., 1991 ). Growth hormone significantly improves survival of rats infected with S. typhimurium and is as effective as IFN-y in preventing death in these rats. One mechanism for growth hormone and IFN-7 promoting rat survival is that both these substances increase the bactericidal activity of macrophages (Edwards et al., 1991, 1992a; Fu et al., 1994). Growth hormone also primes porcine macrophages (Edwards et al., 1988 ) and porcine neutrophils (Fu et al., 1991 ) to secrete superoxide anion, and maximal levels of superoxide anion are comparable to those elicited by porcine IFN-y. Although porcine growth hormone primes porcine neutrophils for superoxide anion secretion and bovine growth hormone primes bovine neutrophils, neither protein is active on human neutrophils (Fu et al., 1992). This was one of the lines of evidence which led us to conclude that human growth hormone actually uses the prolactin receptor to prime human neutrophils for superoxide anion secretion. Leukocytes are now known to synthesize proteins that were once thought to be secreted only by cells of the neuroendocrine system. For example, rat mononuclear cells synthesize and secrete growth hormone (Weigent and Blalock, 1991 ), and we have shown that prolactin is synthesized by human leukocytes (Sabharwal et al., 1992). Furthermore, the 70 amino acid peptide that is induced by growth hormone, IGF-I, is also synthesized by differentiating macrophages (Arkins et al., 1993; Biragyn et al., 1994; Liu et al., 1994). IGF-I primes porcine macrophages (Edwards et al., 1992b) and neutrophils (Fu et al., 1991 ) for secretion of superoxide anion, and substantial evidence suggests IGF-I plays a key role in myeloid cell differentiation as well as T cell differentiation and activation (reviewed by Kelley et al., 1993b). A recent series of peer-reviewed articles describes the roles of IGF-I, growth hormone and prolactin in regulating a variety of immune events (Kelley et al., 1992). The pituitary gland not only secretes hormones which augment certain immune events, but it also secretes hormones that can lead to immunosuppression.
162
K. HI..Kelley et aL / Veterinary Immunology and lmmunopathology 43 (1994) 15 7-165
Injection of antigen has been known for 20 years to increase plasma levels of glucocorticoids. This observation may explain the well-known observation of antigenic competition, where the immune response to a naive antigen administered several days after another antigen is greatly impaired. The rise in glucocorticoids induced by the first antigen suppresses the response to the second one. Several antigens are now known to induce the synthesis of IL-1, which has been repeatedly shown to augment plasma glucocorticoids when given either i.p. or i.c.v. (reviewed by Wick et al., 1992 ). IL-1 increases corticotropin-releasing hormone (CRH) from the medial basal hypothalamus (Spinedi et al., 1992), which in turn causes release of adrenocorticotropin (ACTH) from the adenohypophysis. The subsequent elevation in adrenal-derived glucocorticoids following activation of the immune system probably acts to dampen the clonal expansion of proliferating T and B lymphocytes, and perhaps prevent the development of forbidden clones directed toward self. In support of this hypothesis, animals that are genetically defective in responding to antigen by elevating plasma glucocorticoids are more susceptible to a variety of autoimmune diseases, including streptococcal cell wall arthritis, experimental allergic encephalomyelitis and autoimmune thyroiditis (reviewed by Sternberg and Wilder, 1989). It is not yet known whether plasma cortisol is elevated in domestic pigs following exposure to inflammatory stimuli. Elevation in plasma cortisol in vivo leads to involution of the thymus gland and a reduction in delayed-type hypersensitivity reactions in pigs (Westly and Kelley, 1984). Shipping pigs elevates plasma cortisol and suppresses NK activity, but only in behaviorally submissive pigs (McGlone et al., 1993 ). Similarly, injection of CRH into the lateral ventricle of the brain of pigs elevates plasma ACTH and leads to a 50% reduction in proliferation of PBMC incubated with concanavalin A (Con A; Johnson et al., 1994b), whereas i.v. injection of ACTH also inhibits proliferative responses of porcine PBMC and suppresses IL-2 synthesis (Klemcke et al., 1990). However, ACTH injections have recently been shown to stimulate porcine NK activity, perhaps by indirectly suppressing brain levels of CRH (McGlone et al., 1991). This interpretation is supported by findings in rats where administration of CRH directly into the brain consistently reduces both NK cytotoxicity and T cell proliferation (Irwin et al., 1990; Saperstein et al., 1992). However, this immunosuppression is not mediated by CRH-induced increases in corticosterone because similar effects occur in adrenalectomized rats. These data suggest that CRH itself is an important immunomodulator in the CNS, which is consistent with the expression of receptors for CRH on leukocytes and the direct inflammatory properties of CRH in the periphery (Karalis et al., 1991 ). In experiments conducted in vitro, physiologic concentrations of cortisol inhibit synthesis of TNF-o~ by porcine alveolar macrophages (Dunham et al., 1990) and suppress proliferation of activated porcine thymocytes, splenocytes and PBMC, even though cortisol is only mildly cytolytic (Westly and Kelley, 1984). These effects are probably mediated by specific receptors, because porcine splenocytes express around 2300 glucocorticoid receptors per cell with a KD of 3 nM (Westly and Kelley, 1987 ). The number of glucocorticoid receptors doubles 48 h
K.I~ Kelleyetal./Veterinarylmmunologyandlmmunopathology43(1994) 157-165
163
after addition of Con A, which probably makes these T lymphoblasts more susceptible to the inhibitory effects of glucocorticoids. Glucocorticoids are well known to induce apoptosis in a number of cell types. Therefore, expression of glucocorticoid receptors on specific double positive or double negative cells in the thymus cortex may be critical for normal differentiation of T lymphocytes. One of the most newly recognized anti-inflammatory properties of glucocorticoids is to inhibit the expression of ELAM-1 and ICAMI by activated human endothelial cells (Cronstein et al., 1992 ), but it is unknown if similar events occur in the pig. This is a likely possibility because glucocorticoids consistently cause neutrophilia in pigs (e.g., McGlone et al., 1991 ).
4. Conclusions Leukocytes have 'learned' that they are more effective in controlling inflammation if they employ other physiologic systems. The host's response to infection involves not only the immune system but the entire organism. To ensure coordination of immune responses with physiology and metabolism, leukocytes have developed a signaling network which permits communication not only between cells of the immune system but also among the immune, neuroendocrine and central nervous systems. Inflammatory cytokines, including IL- 1, TNF-a and IL6, not only induce a hepatic acute phase response but also are key molecules that signal the brain that an immune response has been initiated in the periphery. The brain activates homeostatic reactions, such as fever, somnia, anorexia and inactivity, and triggers various responses from the pituitary gland aimed at regulating the activities of both lymphoid and myeloid cells.
Acknowledgment Preparation of this manuscript was supported in part by grants from the National Institutes of Health AG06246 and the United States Department of Agriculture 92-37206-7777.
References S. Arkins, N. Rebeiz, A. Biragyn, D.L. Reese, K.W. Kelley (1993), Endocinology 133, 2334. E. Ban, F. Haour, R. Lenstra (1992), Cytokine 4, 48. E. Ban, G. Milon, N. Prudhomme, G. Fillion, F. Haour ( 1991 ), Neuroscience 43, 21. J.F. Bazan ( 1991 ), Neuron 7, 197. E.N. Benveniste (1992), Am. J. Physiol. 263 (Cell Physiol. 32), C1. A. Biragyn, S. ,Arkins, K.W. Kelley (1993), J. Immunol. Methods 168, 235. J.E. Blalock (1984), J. Immunol. 132, 1067. F. Blecha ( 1991 ), J. Dairy Sci. 74, 328. S.L. Brown, L.R. Smith, J.E. Blalock ( 1987 ), J. Immunol. 139, 3181.
164
K. W. Kelley et al. / Veterinary Immunology and Immunopathology 43 (1994) 15 7-165
D.B. Constam et al. (1992), J. Immunol. 148, 1404. L.J. Cornfield and M.A. Sills ( 1991 ), Eur. J. Pharmacol. 202, 113. B.N. Cronstein, S.C. Kimmel, R.I. Levin, F. Martiniuk, G. Weissmann ( 1992 ), Proc. Natl. Acad. Sci. USA 89, 9991. H.F. Cserr and P.M. Knopf ( 1992 ), Immunol. Today 13,507. E.T. Cunningham et al. (1992), J. Neurosci. 12, 1101. D.A. Cunha, J.A. Jefferson, R.W. Jackson, L. Vitkovic (1993), J. Neuroimmunol. 42, 71. R. Dantzer and K.W. Kelley (1989), Life Sci. 44, 1995. D.M. Dunham, S. Arkins, C.K. Edwards III, R. Dantzer, K.W. Kelley (1990), J. Leukocyte Biol. 48, 473. C.K. Edwards, III et al. (1992a), Infect. Immun. 60, 2514. C.K. Edwards, III et al. (1992b), Cell. Mol. Neurobiol. 12, 499. C.K. Edwards, III, S.M. Ghiasuddin, J.M. Schepper, L.M. Yunger, K.W. Kelley ( 1988 ), Science 239, 769. C.K. Edwards, III, L.M. Yunger, R.M. Lorence, R. Dantzer, K.W. Kelley (1991), Proc. Natl. Acad. Sci. USA 88, 2274. M. Freidin, M.V.L. Bennett, J.A. Kessler (1992), Proc. Natl. Acad. Sci. USA 89, 10440. Y.K. Fu, S. Arkins, B.S. Wang, K.W. Kelley ( 1991 ), J. Immunol. 146, 1602. Y.K. Fu et al. ( 1992 ), J. Clin. Invest. 89, 451. Y.K. Fu, S. Arkins, Y.M. Li, R. Dantzer, K.W. Kelley (1994), Infect. lmmun. 62, 1. W.F. Hickey and H. Kimura (1988), Science 239,290. W.F. Hickey, B.L. Hsu, H. Kimura ( 1991 ), J. Neurosci. Res. 28,254. M. Irwin, W. Vale, C. Rivier (1990), Endocrinology 126, 2837. R.W. Johnson, S.E. Curtis, R. Dantzer, K.W. Kelley (1993a), Physiol. Behav. 53, 127. R.W. Johnson, S.E. Curtis, R. Dantzer, J.M. Bahr, K.W. Kelley (1993b), Physiol. Behav. 53, 343. R.W. Johnson, E. van Borell (1994a), J. Anim. Sci. 72,309. R.W. Johnson, E. van Borell, L.L. Anderson, L.D. Kojic, J.E. Cunnick (1994b), Encinology 135 in press. E. Joly, L. Mucke, M.B.A. Oldstone ( 1991 ), Science 253, 1283. K. Karalis et al. ( 1991 ), Science 254, 421. K.W. Kelley and R. Dantzer (1990), Adv. Vet. Sci. Comp. Med. 35, 283. K.W. Kelley, S. Arkins, Y.M. Li ( 1992 ), Brain, Behav. Immun. 6, 317. K.W. Kelley, S. Kent, R. Dantzer (1993a), in Growth of the Pig, C.R. Hollis, Ed. (CAB International, Wallingford, UK), pp. 119-132. K.W. Kelley, S. Arkins, Y.M. Li (1993b), in Growth Hormone II: Basic and Clinical Aspects, B.B. Bercu and R.R. Walker, Eds. (Springer, Heidelberg, Germany), in press. S. Kent, R.M. Bluth6, K.W. Kelley, R. Dantzer (1992a), Trends Pharmacol. Sci. 13, 24. S. Kent et al. (1992b), Proc. Natl. Acad. Sci. USA 89, 9117. K. Kinouchi, G. Brown, G. Pasternak, D.B. Donner ( 1991 ), Biochem. Biophys. Res. Commun. 181, 1532. H.G. Klemcke, F. Blecha, J.A. Nienaber (1990), Proc. Soc. Exp. Biol. Med. 195, 100. S.C. Lee et al. ( 1993 ), J. Immunol. 150, 594. Q. Liu et al. (1994) Neuroimmunomodulation 1, 33. J.J. McGlone, E.A. Lumpkin, R.L. Norman (1991), Endocrinology 129, 1653. J.J. McGlone, R.I. Nicholson, J.M. Hellman, D.N. Herzog (1993), J. Anim. Sci. 71, 1441. M.F. Mehler, R. Rozenthal, I.M. Dougherty, D.C. Spray, J.A. Kessler ( 1993 ), Nature 362, 62. T.W. Molitor et al. (1992a), J. Leukocyte Biol. 51, 124. T.W. Molitor et al. (1992b), J. Pharmacol. Exp. Therap. 260, 581. J.J. McGlone et al. (1993), J. Anim. Sci. in press. M.C. Morganti-Kossmann, T. Kossmann, S.M. Wahl ( 1992 ), Trends Pharmacol. Sci. 13,286. P. Parnet et al. ( 1993 ), J. Neuroendocrinology, 5, 213. P. Parnet et al. (1994), Molecular Brain Research, in press. P.H. Patterson and H. Nawa (1993), Cell 72/Neuron 10 (Suppl.), 3, 123.
K. W. Kelley et al. / Veterinary lmmunology and lmmunopathology 43 (1994) 157-165
165
P. Sabharwal et al. (1992), Proc. Natl. Acad. Sci. USA 89, 7713. A. Saperstein et al. (1992), Endocrinology 130, 152. B. Schobitz, D.A.M. Voorhuis, E.R. DeKloet (1992), Neurosci. Lett. 136, 189. E. Spinedi, R. Hadid, T. Daneva, R.C. Gaillard ( 1992 ), Neuroendocrinology 56, 46. E.M. Stemberg and R.L. Wilder (1989), Prog. Neuroendocrinimmunology 2, 102. J.L. Tchelingerian, J. Quinonero, J. Booss, C. Jacque (1993), Neuron 10, 213. A.M. van Dam, M. Brouns, S. Louisse, F. Berkenbosch (1992), Brain Res. 588,291. D.A. Weigent and J.E. Blalock ( 1991 ), Cell. Immunol. 135, 55. H.J. Westly and K.W. Kelley (1984), Proc. Soc. Exp. Biol. Med. 177, 156. H.J. Westly and K.W. Kelley (1987), Proc. Soc. Exp. Biol. Med. 185, 211. G. Wick, Y.H. Hu, J. Gruber (1992), Trends Endocrinol. Metab. 3, 141. B.M.R.N.J. Woloski, E.M. Smith, W.J. Meyer III, G.M. Fuller, J.E. Blalock (1985), Science 230, 1035. S. Yamada, M. DePasquale, S. Patlak, H.F. Cserr ( 1991 ), Am. J. Physiol. 26 l, H 1197.
Veterinary immunology .
E LS E VIE R
and
Veterinary Immunology and Immunopathology 43 (1994) 157-165
immunopathology
Immunology discovers physiology Keith W. Kelley a'*, Rodney W. J o h n s o n b, R o b e r t D a n t z e r c aLaboratorv of lmmunophysiology, 207 Plant and Animal Bioteehnology Laboratory, Department of Animal Sciences, University oflllinois, 1201 West Gregory Drive, Urbana, IL 61801, USA b323 Animal Sciences Laboratory, Department of Animal Sciences, University oflllinois, 1201 West Gregory Drive, Urbana, IL 61801, USA CLaboratory of lntegrative Neurobiology, INRA-INSERM, Rue Camille Saint-Sa£;ns 33077 Bordeaux Cedex, France
Abstract
Not so long ago, it was believed that the brain is totally devoid of immunologic reactions, that cytokines derived from activated leukocytes serve only as communication molecules between leukocytes and that the immune system is regulated solely by intrinsic mechanisms. One by one, these old-time, traditional views have fallen by the wayside as neuroscientists, endocrinologists and pharmacologists have begun to explore immunology. The old view was that the immune system is autonomous because it neither affects nor is it affected by other physiologic systems. The new view is that cells of the immune system are inextricably linked with other physiological systems, including the neuroendocrine, cardiovascular, reproductive and central nervous systems (CNS). Changes in one system evoke changes in the other, and it is likely that communication loops have evolved between cells of the immune system and those of other tissues to coordinate and regulate functional activities aimed at preserving homeostasis during inflammation. The integrated view of immunophysiologists that cells of the immune system interact with the entire body, rather than existing as a separate physiologic system that operates autonomously, should help to unravel a number of mysteries in immunoregulation, such as the well-recognized redundant and pleiotropic properties of cytokines. Unfortunately, very few of these ideas have been incorporated into studying immunity of domestic animals. A complete understanding of immunobiology will be achieved only after this new field of immunophysiology is integrated into current immunological thinking. The purpose of this short article is to describe new discoveries which provide insights into how leukocytes discriminate between self and non-self by enlisting the aid of the neuroendocrine and CNS and to document what is known about the immunophysiology of pigs. *Corresponding author. Tel. (217) 333-5141; fax (217) 244-5617. 0165-2427/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0165-2427 ( 94 ) 06020-Z
158
K. V~ Kelley et al. / Veterinary Immunology and Immunopathology 43 (1994) 15 7-165
1. Introduction: the new view of the brain as an immunoprivileged site
The CNS has been considered to be immunologically privileged because it: (a) lacks lymphatic drainage; (b) has relatively impermeable, non-fenestrated cerebral capillary endothelium forming tight junctions (no gap between endothelial cells). The blood-brain barrier is actually a combination of several physical barriers, including the close juxtaposition of endothelial cells, the abundance of astrocytes in the surrounding cerebral capillaries and the small size of the extracellular space; (c) displays very little expression of either class I and II major histocompatibility complex (MHC) molecules; (d) readily accepts allografts. However, this traditional view of the brain as an immunoprivileged site is changing. First, antigens injected into the brain can be found in deep cervical lymph (Yamada et al., 1991 ), apparently escaping the CNS via the cranial nerves, which indicate that there may be substantial traffic from the CNS to lymph. Second, T lymphoblasts do cross the blood-brain barrier, probably by secreting specific endoglycosidases, and do so regardless of MHC phenotype or type of activation signal (Hickey et al., 1991 ). Third, large molecules such as cytokines communicate with the CNS, as demonstrated by i.p. injections of IL-1 causing an elevation in hypothalamic set-point leading to fever. Cytokines could gain access to the CNS via several mechanisms: (a) through circumventricular organs of the brain (which lack a true blood-brain barrier); (b) using glial cells to send messages via prostaglandin E2 to various parts of the CNS; (c) utilizing specific transporters in the cerebral vascular endothelium; or (d) acting indirectly by inducing their own synthesis within the CNS (discussed by Kent et al., 1992a). Finally, perivascular microglia, which develop from the bone marrow and are considered to be a subtype of macrophages in the CNS, can be readily induced to express class II MHC and function as important antigen-presenting cells in experimental allergic encephalomyelitis (Hickey and Kimura, 1988 ). Interferon-y (IFN-7) induces expression of class I MHC molecules on microglia, astrocytes and oligodendrocytes, and class II proteins on microglia and astrocytes (reviewed by Benveniste, 1992). Encephalitic viruses such as lymphocytic choriomeningitis virus persist because neurons lack sufficient expression of class I MHC molecules to effectively engage virus-specific cytotoxic T lymphocytes (Joly et al., 1991 ). These findings are consistent with the new view that immune privilege is an active rather than a passive process that is controlled, in part, by specific suppression of inflammatory responses in the CNS (Cserr and Knopf, 1992). Transforming growth factors ill, if2 and f13 are expected to play important roles in this regard because they are potent suppressive cytokines and they are secreted in both the latent and active forms by microglia and astrocytes (Constam et al., 1992; Cunha et al., 1993). Synthesis of TNF-o~ by lipopolysaccharide (LPS)stimulated porcine alveolar macrophages is significantly suppressed by nanomolar concentrations of TGF-fl2 (Dunham et al., 1990). Expression of class II MHC proteins on IFN-7 treated human microglia is potently down-regulated by colony
K. W. Kelley et al. / Veterinary lmmunology and lmmunopathology 43 (1994) 157-165
159
stimulating factor-I (CSF-1; Lee et al., 1993), which provides another means of suppressing immunologic reactions in the CNS. Leukemia inhibitory factor (LIF)/cholinergic differentiation factor (CDF) is also an important CNS-modulating cytokine (reviewed by Patterson and Nawa, 1993 ). 2. Cytokines and their receptors in the central nervous system Human patients injected with recombinant cytokines often experience chills and have neurological symptoms such as memory loss, drowsiness, weakness and motivational defects (reviewed by Dantzer and Kelley, 1989; Kelley and Dantzer, 1990; Fig. 1 ). Fever, anorexia, sleepiness and lethargy also occur in food animals injected peripherally with recombinant cytokines (Blecha, 1991; F. Blecha, perCENTRAL EFFECTS L" [
Activate Pituitary-Adrenal Axis Increase Thermogenesis Increase Body Temperature Increase Slow-Wave Sleep Sickness Behavior Decrease Body Care Aclwities | Malaise ,~ Decrease Locomotor Activlly Decrease Social Exploration Decrease Food Intake XDecrease Food Motivated BehavLor
._1 < rr I-Z LU
?
4,
LIVER
_1 < rr UJ T
Release of Acute Phase Proteins
JOINTS Pain
13._
cr LU 12.
Immune
Activators
T
GASTROINTESTINAL TRACT Inhibit Gastric Emptying Inhibit Gastric Acid Secretion
t
Hormonal I Effects
.
Cytokines ,~~._ ~,C:~,,w ot ~
I
Autocrine and Paracrine Effects
Immune Ceils
Fig. 1. Proinflammatory cytokines act centrally in the brain. Cytokines amplify not only the immune response in regional lymph nodes but also inform the central nervous system the immune system has been activated. The brain is involved in coordinating a variety of systemic responses, such as inactivity, anorexia, somnia and fever, that promote homeostasis following infection (from Kent et al., 1992a).
160
K. W, Kelley et al. / Veterinary lmmunology and lmmunopathology 43 (1994) 157-165
sonal communication, 1993 ). Since proinflammatory cytokines such as IL-1 and TNF can cause all of these symptoms, we and others have proposed that cytokines act as hormones to inform the CNS that a foreign agent has gained access to the body (Blalock, 1984; Kent et al., 1992a). Although the mechanisms by which peripheral cytokines cause animals to experience clinical signs of sickness are just now beginning to be investigated, we recently used a specific antagonist of the type I IL-1 receptor to show that induction of fever does not share the same receptor mechanisms as those responsible for the decrease in food-motivation and in social activities which are characteristic of sickness (Kent et al., 1992b). Endotoxin induces the synthesis of IL-1 by macrophages in the meninges, choroid plexus, perivascular cells and ramified microglia of the rat brain (Van Dam et al., 1992) and causes a number of behavioral abnormalities in pigs and birds (Johnson et al., 1993a,b; 1994a,b). Trauma-induced IL- I a and T N F a protein (Tchelingerian et al., 1993 ), as well as LPS-induced transcripts for IL- 1o~and ILl fl (Ban et al., 1992), are distributed in specific areas of the brain, suggesting discrete functions for these molecules in the CNS. Sympathetic neurons synthesize and secrete IL-lfl (Freidin et al., 1992). Indeed, a number of cytokines or their transcripts can be detected in the CNS constitutively or during various disease states, including IL-1, IL-2, IL-3, IL-6, CSF-1, IFN-a, IFN-fl, IFN-7, TNF-o~ and TGF-fl (reviewed by Benveniste, 1992), and these cytokines may be key players in both normal and pathogenic states, including debilitating CNS diseases such as AIDS (Morganti-Kossmann et al., 1992). It is particularly interesting that neurotrophic factors such as CDF/LIF, ciliary neurotrophic factor and oncostatin M share significant structural motifs with IL-6 and G-CSF (Bazan, 1991 ). Indeed, the striking similarities between differentiation of cells within the hematopoietic and nervous systems suggest that cytokines evolved as key regulators in both systems (Patterson and Nawa, 1993 ), and that these proteins could be particularly important for differentiation of the newly discovered neuronal stem cells (Mehler et al., 1993 ). Specific receptors for IL-1 have been identified in the hippocampus and dendate gyrus of the mouse brain (Ban et al., 1991; Cunningham et al., 1992), and we have partially cloned both the type I and II IL-1 receptors from the murine adenohypophysis and brain (Parnet et al., 1993; 1994). The pituitary gland also responds to IL-1, IL-2 and IL-6 (Woloski et al., 1985; Brown et al., 1987 ). Receptors for TNF (Kinouchi et al., 1991 ) and IL-6 (Cornfield and Sills, 1991 ; Schobitz et al., 1992 ) are found in the brain, but the specific cell types expressing these receptors are as yet unknown. It is interesting that both the brain and leukocytes express receptors for exogenous (morphine) and endogenous (fl-endorphin) opiates. It appears that porcine natural killer (NK) and phagocytic cells possess naloxone-sensitive opiate receptors because methadone suppresses free radical and NK cytolytic activity in vitro (Molitor et al., 1992a), whereas morphine impairs porcine contact sensitivity reactions in vivo (Molitor et al., 1992b). It has long been known that sick animals do not eat well, and we (Kelley et al., 1993a) have proposed that an immunological mechanism at least partially explains this phenomenon because: (a) sick animals are anorectic; (b) specific re-
K. W. Kelley et al. / Veterinary Immunology and Immunopathology 43 (1994) 15 7-165
161
ceptors for cytokines are found in the CNS; (c) several bacterial and viral pathogens induce synthesis and secretion of cytokines by myeloid cells; (d) injection of recombinant cytokines cause animals to stop eating; (e) the anorectic effect of at least one cytokine (IL-1 ) is blocked by a specific receptor antagonist. We think that cytokines are somehow involved in normal brain functions that control economically important traits of farm animals, such as food intake.
3. The pituitary gland as a link between the immune and central nervous systems
Relationships between the immune system and the brain are truly bidirectional in the sense that not only products from the immune system affect the brain, but also hormones and neurotransmitters modulate a number of immune events. This is particulary true for somatotropin (growth hormone), a hormone secreted by the anterior pituitary gland. In the absence of a pituitary gland (hypophysectomy), we have shown that rats infected with Salmonella typhimurium die much more rapidly than their controls with a hypophysis (Edwards et al., 1991 ). Growth hormone significantly improves survival of rats infected with S. typhimurium and is as effective as IFN-y in preventing death in these rats. One mechanism for growth hormone and IFN-7 promoting rat survival is that both these substances increase the bactericidal activity of macrophages (Edwards et al., 1991, 1992a; Fu et al., 1994). Growth hormone also primes porcine macrophages (Edwards et al., 1988 ) and porcine neutrophils (Fu et al., 1991 ) to secrete superoxide anion, and maximal levels of superoxide anion are comparable to those elicited by porcine IFN-y. Although porcine growth hormone primes porcine neutrophils for superoxide anion secretion and bovine growth hormone primes bovine neutrophils, neither protein is active on human neutrophils (Fu et al., 1992). This was one of the lines of evidence which led us to conclude that human growth hormone actually uses the prolactin receptor to prime human neutrophils for superoxide anion secretion. Leukocytes are now known to synthesize proteins that were once thought to be secreted only by cells of the neuroendocrine system. For example, rat mononuclear cells synthesize and secrete growth hormone (Weigent and Blalock, 1991 ), and we have shown that prolactin is synthesized by human leukocytes (Sabharwal et al., 1992). Furthermore, the 70 amino acid peptide that is induced by growth hormone, IGF-I, is also synthesized by differentiating macrophages (Arkins et al., 1993; Biragyn et al., 1994; Liu et al., 1994). IGF-I primes porcine macrophages (Edwards et al., 1992b) and neutrophils (Fu et al., 1991 ) for secretion of superoxide anion, and substantial evidence suggests IGF-I plays a key role in myeloid cell differentiation as well as T cell differentiation and activation (reviewed by Kelley et al., 1993b). A recent series of peer-reviewed articles describes the roles of IGF-I, growth hormone and prolactin in regulating a variety of immune events (Kelley et al., 1992). The pituitary gland not only secretes hormones which augment certain immune events, but it also secretes hormones that can lead to immunosuppression.
162
K. HI..Kelley et aL / Veterinary Immunology and lmmunopathology 43 (1994) 15 7-165
Injection of antigen has been known for 20 years to increase plasma levels of glucocorticoids. This observation may explain the well-known observation of antigenic competition, where the immune response to a naive antigen administered several days after another antigen is greatly impaired. The rise in glucocorticoids induced by the first antigen suppresses the response to the second one. Several antigens are now known to induce the synthesis of IL-1, which has been repeatedly shown to augment plasma glucocorticoids when given either i.p. or i.c.v. (reviewed by Wick et al., 1992 ). IL-1 increases corticotropin-releasing hormone (CRH) from the medial basal hypothalamus (Spinedi et al., 1992), which in turn causes release of adrenocorticotropin (ACTH) from the adenohypophysis. The subsequent elevation in adrenal-derived glucocorticoids following activation of the immune system probably acts to dampen the clonal expansion of proliferating T and B lymphocytes, and perhaps prevent the development of forbidden clones directed toward self. In support of this hypothesis, animals that are genetically defective in responding to antigen by elevating plasma glucocorticoids are more susceptible to a variety of autoimmune diseases, including streptococcal cell wall arthritis, experimental allergic encephalomyelitis and autoimmune thyroiditis (reviewed by Sternberg and Wilder, 1989). It is not yet known whether plasma cortisol is elevated in domestic pigs following exposure to inflammatory stimuli. Elevation in plasma cortisol in vivo leads to involution of the thymus gland and a reduction in delayed-type hypersensitivity reactions in pigs (Westly and Kelley, 1984). Shipping pigs elevates plasma cortisol and suppresses NK activity, but only in behaviorally submissive pigs (McGlone et al., 1993 ). Similarly, injection of CRH into the lateral ventricle of the brain of pigs elevates plasma ACTH and leads to a 50% reduction in proliferation of PBMC incubated with concanavalin A (Con A; Johnson et al., 1994b), whereas i.v. injection of ACTH also inhibits proliferative responses of porcine PBMC and suppresses IL-2 synthesis (Klemcke et al., 1990). However, ACTH injections have recently been shown to stimulate porcine NK activity, perhaps by indirectly suppressing brain levels of CRH (McGlone et al., 1991). This interpretation is supported by findings in rats where administration of CRH directly into the brain consistently reduces both NK cytotoxicity and T cell proliferation (Irwin et al., 1990; Saperstein et al., 1992). However, this immunosuppression is not mediated by CRH-induced increases in corticosterone because similar effects occur in adrenalectomized rats. These data suggest that CRH itself is an important immunomodulator in the CNS, which is consistent with the expression of receptors for CRH on leukocytes and the direct inflammatory properties of CRH in the periphery (Karalis et al., 1991 ). In experiments conducted in vitro, physiologic concentrations of cortisol inhibit synthesis of TNF-o~ by porcine alveolar macrophages (Dunham et al., 1990) and suppress proliferation of activated porcine thymocytes, splenocytes and PBMC, even though cortisol is only mildly cytolytic (Westly and Kelley, 1984). These effects are probably mediated by specific receptors, because porcine splenocytes express around 2300 glucocorticoid receptors per cell with a KD of 3 nM (Westly and Kelley, 1987 ). The number of glucocorticoid receptors doubles 48 h
K.I~ Kelleyetal./Veterinarylmmunologyandlmmunopathology43(1994) 157-165
163
after addition of Con A, which probably makes these T lymphoblasts more susceptible to the inhibitory effects of glucocorticoids. Glucocorticoids are well known to induce apoptosis in a number of cell types. Therefore, expression of glucocorticoid receptors on specific double positive or double negative cells in the thymus cortex may be critical for normal differentiation of T lymphocytes. One of the most newly recognized anti-inflammatory properties of glucocorticoids is to inhibit the expression of ELAM-1 and ICAMI by activated human endothelial cells (Cronstein et al., 1992 ), but it is unknown if similar events occur in the pig. This is a likely possibility because glucocorticoids consistently cause neutrophilia in pigs (e.g., McGlone et al., 1991 ).
4. Conclusions Leukocytes have 'learned' that they are more effective in controlling inflammation if they employ other physiologic systems. The host's response to infection involves not only the immune system but the entire organism. To ensure coordination of immune responses with physiology and metabolism, leukocytes have developed a signaling network which permits communication not only between cells of the immune system but also among the immune, neuroendocrine and central nervous systems. Inflammatory cytokines, including IL- 1, TNF-a and IL6, not only induce a hepatic acute phase response but also are key molecules that signal the brain that an immune response has been initiated in the periphery. The brain activates homeostatic reactions, such as fever, somnia, anorexia and inactivity, and triggers various responses from the pituitary gland aimed at regulating the activities of both lymphoid and myeloid cells.
Acknowledgment Preparation of this manuscript was supported in part by grants from the National Institutes of Health AG06246 and the United States Department of Agriculture 92-37206-7777.
References S. Arkins, N. Rebeiz, A. Biragyn, D.L. Reese, K.W. Kelley (1993), Endocinology 133, 2334. E. Ban, F. Haour, R. Lenstra (1992), Cytokine 4, 48. E. Ban, G. Milon, N. Prudhomme, G. Fillion, F. Haour ( 1991 ), Neuroscience 43, 21. J.F. Bazan ( 1991 ), Neuron 7, 197. E.N. Benveniste (1992), Am. J. Physiol. 263 (Cell Physiol. 32), C1. A. Biragyn, S. ,Arkins, K.W. Kelley (1993), J. Immunol. Methods 168, 235. J.E. Blalock (1984), J. Immunol. 132, 1067. F. Blecha ( 1991 ), J. Dairy Sci. 74, 328. S.L. Brown, L.R. Smith, J.E. Blalock ( 1987 ), J. Immunol. 139, 3181.
164
K. W. Kelley et al. / Veterinary Immunology and Immunopathology 43 (1994) 15 7-165
D.B. Constam et al. (1992), J. Immunol. 148, 1404. L.J. Cornfield and M.A. Sills ( 1991 ), Eur. J. Pharmacol. 202, 113. B.N. Cronstein, S.C. Kimmel, R.I. Levin, F. Martiniuk, G. Weissmann ( 1992 ), Proc. Natl. Acad. Sci. USA 89, 9991. H.F. Cserr and P.M. Knopf ( 1992 ), Immunol. Today 13,507. E.T. Cunningham et al. (1992), J. Neurosci. 12, 1101. D.A. Cunha, J.A. Jefferson, R.W. Jackson, L. Vitkovic (1993), J. Neuroimmunol. 42, 71. R. Dantzer and K.W. Kelley (1989), Life Sci. 44, 1995. D.M. Dunham, S. Arkins, C.K. Edwards III, R. Dantzer, K.W. Kelley (1990), J. Leukocyte Biol. 48, 473. C.K. Edwards, III et al. (1992a), Infect. Immun. 60, 2514. C.K. Edwards, III et al. (1992b), Cell. Mol. Neurobiol. 12, 499. C.K. Edwards, III, S.M. Ghiasuddin, J.M. Schepper, L.M. Yunger, K.W. Kelley ( 1988 ), Science 239, 769. C.K. Edwards, III, L.M. Yunger, R.M. Lorence, R. Dantzer, K.W. Kelley (1991), Proc. Natl. Acad. Sci. USA 88, 2274. M. Freidin, M.V.L. Bennett, J.A. Kessler (1992), Proc. Natl. Acad. Sci. USA 89, 10440. Y.K. Fu, S. Arkins, B.S. Wang, K.W. Kelley ( 1991 ), J. Immunol. 146, 1602. Y.K. Fu et al. ( 1992 ), J. Clin. Invest. 89, 451. Y.K. Fu, S. Arkins, Y.M. Li, R. Dantzer, K.W. Kelley (1994), Infect. lmmun. 62, 1. W.F. Hickey and H. Kimura (1988), Science 239,290. W.F. Hickey, B.L. Hsu, H. Kimura ( 1991 ), J. Neurosci. Res. 28,254. M. Irwin, W. Vale, C. Rivier (1990), Endocrinology 126, 2837. R.W. Johnson, S.E. Curtis, R. Dantzer, K.W. Kelley (1993a), Physiol. Behav. 53, 127. R.W. Johnson, S.E. Curtis, R. Dantzer, J.M. Bahr, K.W. Kelley (1993b), Physiol. Behav. 53, 343. R.W. Johnson, E. van Borell (1994a), J. Anim. Sci. 72,309. R.W. Johnson, E. van Borell, L.L. Anderson, L.D. Kojic, J.E. Cunnick (1994b), Encinology 135 in press. E. Joly, L. Mucke, M.B.A. Oldstone ( 1991 ), Science 253, 1283. K. Karalis et al. ( 1991 ), Science 254, 421. K.W. Kelley and R. Dantzer (1990), Adv. Vet. Sci. Comp. Med. 35, 283. K.W. Kelley, S. Arkins, Y.M. Li ( 1992 ), Brain, Behav. Immun. 6, 317. K.W. Kelley, S. Kent, R. Dantzer (1993a), in Growth of the Pig, C.R. Hollis, Ed. (CAB International, Wallingford, UK), pp. 119-132. K.W. Kelley, S. Arkins, Y.M. Li (1993b), in Growth Hormone II: Basic and Clinical Aspects, B.B. Bercu and R.R. Walker, Eds. (Springer, Heidelberg, Germany), in press. S. Kent, R.M. Bluth6, K.W. Kelley, R. Dantzer (1992a), Trends Pharmacol. Sci. 13, 24. S. Kent et al. (1992b), Proc. Natl. Acad. Sci. USA 89, 9117. K. Kinouchi, G. Brown, G. Pasternak, D.B. Donner ( 1991 ), Biochem. Biophys. Res. Commun. 181, 1532. H.G. Klemcke, F. Blecha, J.A. Nienaber (1990), Proc. Soc. Exp. Biol. Med. 195, 100. S.C. Lee et al. ( 1993 ), J. Immunol. 150, 594. Q. Liu et al. (1994) Neuroimmunomodulation 1, 33. J.J. McGlone, E.A. Lumpkin, R.L. Norman (1991), Endocrinology 129, 1653. J.J. McGlone, R.I. Nicholson, J.M. Hellman, D.N. Herzog (1993), J. Anim. Sci. 71, 1441. M.F. Mehler, R. Rozenthal, I.M. Dougherty, D.C. Spray, J.A. Kessler ( 1993 ), Nature 362, 62. T.W. Molitor et al. (1992a), J. Leukocyte Biol. 51, 124. T.W. Molitor et al. (1992b), J. Pharmacol. Exp. Therap. 260, 581. J.J. McGlone et al. (1993), J. Anim. Sci. in press. M.C. Morganti-Kossmann, T. Kossmann, S.M. Wahl ( 1992 ), Trends Pharmacol. Sci. 13,286. P. Parnet et al. ( 1993 ), J. Neuroendocrinology, 5, 213. P. Parnet et al. (1994), Molecular Brain Research, in press. P.H. Patterson and H. Nawa (1993), Cell 72/Neuron 10 (Suppl.), 3, 123.
K. W. Kelley et al. / Veterinary lmmunology and lmmunopathology 43 (1994) 157-165
165
P. Sabharwal et al. (1992), Proc. Natl. Acad. Sci. USA 89, 7713. A. Saperstein et al. (1992), Endocrinology 130, 152. B. Schobitz, D.A.M. Voorhuis, E.R. DeKloet (1992), Neurosci. Lett. 136, 189. E. Spinedi, R. Hadid, T. Daneva, R.C. Gaillard ( 1992 ), Neuroendocrinology 56, 46. E.M. Stemberg and R.L. Wilder (1989), Prog. Neuroendocrinimmunology 2, 102. J.L. Tchelingerian, J. Quinonero, J. Booss, C. Jacque (1993), Neuron 10, 213. A.M. van Dam, M. Brouns, S. Louisse, F. Berkenbosch (1992), Brain Res. 588,291. D.A. Weigent and J.E. Blalock ( 1991 ), Cell. Immunol. 135, 55. H.J. Westly and K.W. Kelley (1984), Proc. Soc. Exp. Biol. Med. 177, 156. H.J. Westly and K.W. Kelley (1987), Proc. Soc. Exp. Biol. Med. 185, 211. G. Wick, Y.H. Hu, J. Gruber (1992), Trends Endocrinol. Metab. 3, 141. B.M.R.N.J. Woloski, E.M. Smith, W.J. Meyer III, G.M. Fuller, J.E. Blalock (1985), Science 230, 1035. S. Yamada, M. DePasquale, S. Patlak, H.F. Cserr ( 1991 ), Am. J. Physiol. 26 l, H 1197.