Neural and biochemical mediators of endotoxin and stress-induced c-fos expression in the rat brain

Brain Research Bulletin, Vol. 34, No. 1, pp. 7-14, 1994 Copyright 0 1994 Ebevier ScienceL.td Printed in the USA. All rights reserved

Pergfimon

0361-9230/94$6.00 + .oO

0361-9230(93)EOO34-J

Neural and Biochemical Mediators of Endotoxin and Stress-Induced c-fos Exnression in the Rat Brain 1

WEIHUA WAN,* LISA WETMORE,* CRAIG M. SORENSEN,-f ARNOLD H. GREENBERG* AND DWIGHT M. NANCE*’ *Departments of Pathology and Physiology and The Manitoba Institute of Cell Biology, University of Manitoba, Winnipeg, Manitoba, R3E OW3, Canada, ‘I’OncogeneScience, Inc., Uniondale, NY 11553 Received

8 August

1993; Accepted

30 November

1993

WAN, W., L. WETMORE, C. M. SORENSEN, A. H. GREENBERG AND D. M. NANCE. Neural und b~~~rnica~ mediators of endotoxinand stress-inducedc-fos expression in the rat brain. BRAIN RES BULL 34(l) 7-14, 1994.-We and others bave reported that c-fos protein is induced in the hypothalamus and brain stem of the rat following central and peripheral injections of endotoxin (lipopolysaccharide; LPS). We have now examined possible mechanisms through which LPS induces c-fos protein. The cycloxygenase inhibitor indomethacin and the glutamate NMDA antagonist MK801 inhibited c-fos protein in the paraventricular nucleus (PVN), supraoptic nucleus (SON), and the Al/A2 regions of the brain stem induced by IP or IV injections of LPS (40 pg). The Hl histamine antagonist diphenhydramine, but not the H2 histamine antagonist cimetidine, reduced the amount of c-for labeling. MK801 also attenuated the effects of stress (foot shock) on c-fos protein; however, indomethacin bad no effect on c-fos protein induced by stress. We next examined the importance of visceral afferent innervation on the response to LPS or stress. Subdiaphragmatic vagotomy completely blocked the induction of c-fos protein following IP injections of LPS; however, vagotomy bad a minimal effect on c-fos protein induced in the PVN and SON following IV injections of LPS, but potentiated c-fos induction following foot shock. Thus, prostaglandin synthesis, glutamate release, histamine receptors, and visceral afferents represent functional biochemical and neural pathways through which endotoxin activates c-fos protein in specific autonomic and neuroendocrine regulatory nuclei. Activation of NMDA glutamate receptors may represent a final common pathway for the induction of c-fos protein in the brain induced by botb endotoxin and stress. Lipopolysaccharide c-fos protein Foot shock Indomethacin Histamine antagonist MK801

Paraventricular nucleus

THE immediate early gene, c-fos, is differentially expressed in multiple regions of the brain and spinal cord following different physiological challenges (17,19,30-3258). We and others have demonstrated that c-fos protein is induced selectively in neurons located in the hypothalamus and brain stem of the rat following administration of lipo~lysa~haride (LPS) or foot shock (47,56,64,66). The indu~ion of c-fos has been implicated in diverse processes such as cell differentiation, proliferation, and signal transduction (34,41). Although the precise function of c-fos protein in the CNS remains unclear, the neuronal expression of c-fos protein provides a dynamic view of complex neural regulatory networks that can be analyzed further at the cellular level (31). Lipopolysaccharide (LPS), an endotoxin associated with gram-negative bacteria, is a potent inducer of cytokines such as interleukin-1 (IL-l) and tumor necrosis factor (TNF) (3,70,71), and various biological actions of LPS have been demonstrated to

Supraoptic nucleus

Vagotomy

be secondary to the production of these cytokines (33). Prostaglandins can modify the production of TNF and IL-1 (13,23), and have been shown to mediate some of the effects of these cytokines as well as LPS (6,16,26,43,59). Histamine, another important immune system mediator, has been reported to suppress the production of both IL-l and TNF induced by LPS in vitro and to mediate LPS-induced co~icosterone release via both Hl and H2 histamine receptors (37,51,63). Evidence suggests that the induction of c-fos protein is mediated via calcium entry (54) into neurons with NMDA glutamate receptors (49). MKSOl, a NMDA glutamate receptor antagonist, has been shown to prevent the activation of c-fos protein following brain injury (l&17), cocaine administration (58) as well as light exposure (1). In the present study we have examined whether induction of c-fos protein in the brain following LPS stimulation or stress is dependent on any of the above neural and biochemical mediators.

’ Requests for reprints should be addressed to D. M. Nance, Department of Pathology, University of Manitoba, 770 Barmatyne Ave., Winnipeg, Manitoba, R3E 0W3, Canada.

7

8

WAN ET AL.

METHOD

Subjects Adult (160-510 g) male Sprague-Dawley rats (Charles River, Dorval, Quebec) were maintained in groups of two to three per cage with food and water ad lib in a reversed light room with a 12:12 lighting schedule. Chemicals LPS (E. coli 055:B5), indomethacin, nordihydroguaiaretic acid (NDGA), diphenhydrimine, and cimetidine were purchased from Sigma (StLouis, MO), and MK801 was purchased from Research Biochemicals, Inc. (Natick, MA). Drug Administrations

and LPS injections

Rats were given IP injections of either indomethacin (50 mg/ kg), NDGA (125 mg/kg), diphenhydramine (10 mg/kg), cimetidine (100 mgikg), MK801 (3 mg/kg) or else vehicle. Indomethacin, cimetidine, diphenhydramine, and MK801 were all dissolved in saline, whereas NDGA was dissolved in 50% DMSO. Following a 30 min interval, rats were then given an IP injection of LPS (40 pg/rat in 0.2 ml) or saline and 2 h later overdosed with pentabarbital and perfused with fixative. In additional groups of animals, either Indomethacin, MK801 or saline were administered IP 30 min before IV injections of 40 pg of LPS/O.2 ml via a tail vein. Two hours following the IV injections, the animals were overdosed and perfused with fixative. To examine the effects of these drug treatments upon stressinduced c-fos protein, rats were given IP injections of either indomethacin (50 mg/kg), MK801 (5 mg/kg), or else saline. After a 30 min interval, rats were either placed in a standard rat operant chamber and exposed for 60 min to an intermittent signaled foot shock [5 s of 1.6 mA scrambled shock on a variable interval 2.5 min schedule with each shock being preceeded by a 15 s warning tone; see (65)] or else the animals were placed in a standard rat Plexiglas restrainer for 60 min. Nonstressed controls consisted of animals that were either placed in the operant chamber for 60 min with the shock grid turned off or else left undisturbed in their home cages. At the end of the tests, animals were overdosed and perfused with fixative. Effects of Subdiaphragmatic

Vagotomy

In a final set of experiments, animals were anesthetized with pentabarbital(60 mg/kg) and with the aid of a dissecting microscope, the vagus nerve was surgically sectioned below the diaphragm or else sham surgeries performed. Following a 7 day recovery period, the vagotomized and sham operated animals were either given an IP or IV injection of LPS (40 pg) as outlined above and perfused with fixative 2 h later. Additional groups of vagotomized and sham-operated rats were exposed for 60 min to either intermittent foot shock or restraint stress as outlined above, and the animals overdosed and perfused at the end of the stress tests. Completness of the vagotomies was verified by weighing the stomach following perfusion. Immunocytochemical

Procedure

All animals were perfused transcardially with 100 ml of 1.0% sodium nitrite followed by 400 ml of fresh 4.0% paraformaldehyde in phosphate buffer (pH = 7.3). Brains were postfixed for 2-4 h and then placed into 30% buffered sucrose for l-2 days prior to sectioning. As outlined previously (64) serial 40 pm sections were cut on a freezing microtome and alternate sections were transferred to 24-well culture plates containing 0.01 M

phosphate buffer saline (PBS) and washed for 30 min. Sections were then incubated on a rocker table with a rabbit antibody to c-fos protein (Oncogene Sci., Uniondale, NY) diluted 1:200 in a PBS solution containing 1% normal goat serum and 1% Triton X-100 overnight at room temperature. The following day sections were washed three times in PBS and then incubated in goat antirabbit antibody (1:150; Cappel, Scarborough, Ontario) for 90 min. Sections were again washed three times in PBS and incubated for 90 min in rabbit PAP (1:300; Cappel). Sections were then rinsed in PBS and transferred to plates containing the chromogen diaminobenzidine, d-glucose, and glucose oxidase. The peroxidase reaction was visualized following a 30-45 min incubation. Sections were rinsed, floated onto subbed slides, air dried, cleared in alcohol/xylene, and coverslipped with Permount. Data Analysis Brain sections taken at comparable levels of the paraventricular nucleus (PVN) and supraoptic nucleus (SON) of the hypothalamus of each animal were used to index the number of c-fos-labelled cells for each group. The most heavily labelled single section through the PVN or SON of each animal was chosen for cell counts. All cells labelled in a single section were counted, regardless of the intensity of staining, in a standardized manner through a microscope using a grid reticule. All data were analyzed by an analysis of variance (CLR ANOVA, Houston, TX) and group differences assessed by orthogonal contrast and F-ratios or Duncan’s test (67). RESULTS

The patterns of c-fos immunostaining in the hypothalamus and brain stem were comparable to that described previously for LPS or stress (64,66). Relative to vehicle-injected controls, there was a dramatic increase in the number of c-fos-positive neurons localized in the PVN, SON, arcuate nucleus, and the Al/A2 regions of the brain stem following both IP and IV injections of LPS (Fig. 1). In the stressed animals, additional positive cells were observed in the lateral septal area, amygdala, preoptic area, and lateral hypothalamic region. As illustrated in Fig. 2, IV injections of LPS produced significantly more c-fos-labeled neurons in the PVN and SON than did IP injections, F(1,22) = 5.023, p < 0.03. Pretreatment with the cycloxygenase inhibitor indomethacin (50 mg/kg, IP) significantly, F(1, 22) = 25.154, P < 0.01, reduced the number of cfos-positive neurons in both nuclei following both IP (-77%) and IV (-81%) LPS injections, relative to saline pretreated animals (Fig. 2). Animals treated with indomethacin followed by saline were comparable to saline-treated controls (data not shown). In contrast, pretreatment with the lipoxygenase inhibitor NDGA (125 mg/kg, IP) had no effect on LPS-induced c-fos expression in the brain (data not shown). However, injections of NDGA alone produced numerous c-fos-labeled neurons in the PVN and SON that were indistinguishable from LPS-treated rats. Therefore, the NDGA results were not intepretable. The activation of c-fos protein by NDGA was not due to the vehicle because IP injections of the 50% DMSO solution were without effect. Also shown in Fig. 2, pretreatment with the NMDA antagonist MK801 (3 mgkg, IP) significantly reduced the number of PVN and SON neurons expressing c-fos following both IP (-64%) and IV (-69%) LPS injections, relative to animals pretreated with saline, F(1, 22) = 23.314, p < 0.01. In animals pretreated with the Hl histamine inhibitor diphenhydramine (10 mg/kg, IP) the induction of c-fos protein produced by IP injections of LPS was significantly inhibited, F(2, 22) =

REGULATION OF BRAIN c-fos

FIG. 1. Representative photomicrographs illustrating the effects of IP and IV injections of LPS (40 /lp) and stress (foot shock) on the expression of c-f& protein in the paraventricular nucleus (PVN) of the hypothalamus as well as the effects of pretreatment with indomethacin, MK801 or else subdiaphragmatic vagotomy. Illustrated in A is the PVN of an animal given a sham operation 1 week prior to an IP injection of LPS, and B illustrates the inhibitory effect of a subdiaphragmatic vagotomy on the induction of c- fos protein following an IP injection of LPS. C shows the effects of an IV injection of LPS on c-fos expression in the PVN, whereas D illustrates the minimal effect of vagotomy on c- fosexpression in the PVN following IV injections of LPS. E and F illustrate the inhibitory effect if pretreatment with indomethacin or MK801, respectively, on the expression of c- fos protein in the PVN following an IV iniection of LPS. G shows the expression of c- fos protein in the PVN following 60 min of intermittent foot shock, &d H illustrates the inhibitory kfect of pretreatment with MK801 on the induction of c- fos protein produced by stress. Refer to text for details.

WAN ET AL.

10

IPLPS

l

ative to sham-operated animals, F&8) = 6.325,~ < 0.04. Stomach weights verified the completeness of the vagotomies with the stomach weights of the vagotomized rats being two to three times greater than the sham-operated animals (lo-17 g vs. 3-7 g).

IV LPS

r

*

cf

PVN

•d

SON

-I-

DISCUSSION

We have demonstrated that c-fos protein is induced in the PVN, SON, arcuate nucleus, as well as the Al and A2 regions of the brain stem by systemic injections of LPS. Additionally, an increase in the number of c-fos-positive cells was observed in other brain regions such as the lateral septal area (LSA), bed nucleus of the stria terminals, preoptic area (POA), amygdala, and locus coeruleus following foot shock. Pretreatment with the cycloxygenase inhibitor indomethacin blocked c-fos expression in both the PVN and SON of the rat brain following LPS injection, but had no effect on foot shock-induced c-fos expression. In preliminary studies, we found that indomethacin was ineffective if administered at the same time as the LPS injection, indicating that there is a critical time interval required for it to exert its inhibitory effect on c-fos expression. Indomethacin can alter the action of LPS on cytotoxic activity in vitro (62), and PGE, is involved in controlling the release of TNF from LPS-stimulated macrophages. The dose of indomethacin we utilized has been shown to completely block LPS-induced changes in hypothalamic catecholamine activity (25). In Swiss 3T3 cells, indomethacin has been shown to block PC& release and CAMP accumulation and subsequently reduce c-fos induction in response to bombesin (26), suggesting that the effects of PGE&on CAMP levels contributes to the expression of c-fos protein. On the other hand, the lipoxygenase inhibitor NDGA has been shown to abolish the in vitro induction of c-fos mRNA by TNF (16). Unfortunately, because we found that the lipoxygenase inhibitor NDGA produced a similar number of c-fos immunostained cells in the brain as did LPS, any possible role for lipoxygenase metabolites in mediating the effects of LPS on c-fos expression in the brain remains open. Interestin~y, indomethacin has been shown to prevent the LPS-induced inhibition of gastric secretion (59). This, together with the present data showing that both vagotomy and indomethacin can block LPS-induced c-fos expression, indicates that systemic (IP) LPS requires an intact vagus nerve to stimulate c-fos immunoreactivity in the rat brain via a prostagiandin-dependent mechanism. In addition, cytokines,

FIG. 2. Effects of pretreatment with indomethacin (INDO) or MK801 on the mean (2 SE) number of c-@-labeled neurons in the paraventricular nucleus (PVN) and supraoptic nucleus (SON) of the hypothalamus following either IP or IV injections of 40 pg of LPS. Asterisk (*) indicates significantly different from saline-treated controls and INDOI MKgOl pretreated groups. Double asterisks (**) indicate significantly different from saline-treated controls and ~O~K801 pretreated groups and significantly different from the animals treated with IP LPS. Refer to text for details.

4.278,~ < 0.03, -53%, relative to animals pretreated with saline (Fig. 3). On the other hand, the HZ histamine inhibitor cimetidine (100 mg/kg, IP) had no effect on c-&r expression under the same conditions (Fig. 3). The effects of indomethacin and MK801 on c-fos expression in the hypothalamus following foot shock are shown in Fig. 4. Foot shock produced a significant increase in the number of cfor-labeled neurons in the PVN and SON, relative to the no shock controls, F(1, 18) = 29.115, p < 0.01. In the absence of foot shock, there were no significant group differences or treatment effects, F(2, 18) = 1.324, p < 0.29. Foot shock still produced a significant increase in the number of labeled cells in the PVN and SON following pretreatment with SO mg/kg indomethacin, F(1, 18) = 18.611, p < 0.01; however, increased c-fm labeling was not observed following pretreatment with MK801 (5 mg/kg), F(1, 18) = 0.405, p < 0.53. The activation of c-fos protein in additional brain areas, such as the lateral septai area and amygdala, typically observed following foot shock was also inhibited by MK801 (data not shown). The effects of vagotomy on the induction of c-fos protein in the PVN and SON following IF or IV LPS are illustrated in Fig. 5. Injections (IV) produced si~ific~dy more labeled cells in the hypothal~us than did IP injections, F(1, 11) = 15.761, p < 0.01. Subdiaphragmatic vagotomy completely blocked the activation of c-fos protein in both the PVN and SON following IP injections of LPS, F(1, 11) = 31.635, p < 0.01. Although vagotomy produced a small and statistically significant, F( 1,11) = 6.246,~ < 0.03, decrease in the hypothalamus following IV LPS injections, the ma~itude of the inhibition was modest in comparison to the dramatic effects on IP injections. Sham surgery itself had no effect in comparison to homecaged controls (data not shown). In contrast to LPS, vagotomy failed to block restraint stress or foot shock-induced c-fos expression in any of the brain nuclei examined (Fig. 6). Actually, vagotomy significantly increased the number of labeled neurons in the PVN following foot shock rel-

600

500

2 $

400

2i

300

w 0

0

PVN

.

lZl

SON

-r

I

Hl

H2

200

* 100

0

~6 SAL/SAL

10

SALiLPS

DPHlSAL

DPHiLPS

CMDiSAL

CMDiLPS

GROUPS

FIG. 3. Effects of pretreatment with either a HI (diphenhydramine; DPH) or H2 (cimetidine; CMD) histamine antagonist of the expression of cfos protein in the PVN and SON following IP injections of LPS (40 pg). Asterisk (*) indicates significantly different from appropriate control groups. Refer to text for details.

REGULATION

11

OF BRAIN c-fos 600

MK801

INDO

SALINE

-I-

**

L

1II

II

6

0 SAL-FS

jAL+FS

INDO-FS

INDO+FS

MKSOl-FS

MKBOl+FS

GROUPS

FIG. 4. Effects of pretreatment with either indomethacin (INDO) pression of c-for protein in the PVN and SON in rats following foot shock (+FS) or else animals not exposed to foot shock (-FS). asterisks (**) indicate significantly different from similarly treated foot shock. Refer to text for details.

such as IL-l, are released by macrophages following LPS stimulation, and these cytokines may directly activate primary afferent nociceptor neurons (24). Thus, the release of prostaglandins, as well as histamine (see below), represent primary mediators of the action of endotoxin on peripheral afferents. Because the activational effects of IV injections of LPS on c-fos are also blocked by indomethacin, but not by vagotomy, the action of LPS at more central sites (circumventricular organs) is likely mediated via the release of prostaglandins. It will be important to determine whether central injections of prostaglandin inhibitors can also block the effects of systemic injections of LPS. Because the induction of c-fos in the brain produced by foot shock was unaltered by indomethacin, it is clear that the activational effects of stress are mediated by other pathways. Histamine is released from basophils and mast cells during immediate-type hypersensitivity reactions, and many of its inflammatory effects are mediated by histamine Hl receptors, whereas various immunoregulatory effects are mediated via H2

or MK801 on the ex60 min of intermittent Asterisk (*) and double groups not exposed to

receptors. The regulation of TNF or IL-1 synthesis by histamine in vitro is mediated by both Hl and H2 receptors (3763). The elevation of corticosterone levels produced by LPS is reported to be mediated by the production of histamine in peripheral tissues which acts via Hl histamine receptors (51). Histamine has been shown to stimulate the proliferation of airway smooth muscle and to induce c-fos expression (40). We have observed that IP injections of histamine does stimulate some c-fos expression in the PVN, but the number of c-fos-positive cells in the PVN was less than half of the number produced by LPS (unpublished data). We found that the histamine Hl receptor antagonist, diphenhydramine, reduced LPS-induced c-fos expression in the PVN and SON of the rat brain, but the reduction was somewhat less than that observed with indomethacin. These results suggest that LPSinduced c-fos expression in the brain is at least partially mediated by the peripheral release of histamine, which presumably acts upon Hl receptors. On the other hand, cimetidine, a histamine H2 receptor antagonist, has been reported to significantly enhance LPS-induced synthesis of TNF (63) and to inhibit nonopiate-dependent analgesia induced by stress (18). The present dose

IV LPS RESTRAINT

i

E

200

* 100

0 SHAMIIP-LPS

VAGOTIIP-LPS

SHAM/IV-LPS

VAGOT/lV-LPS

GROUPS

0: SHAM+FS

FIG. 5. Effects of subdiaphragmatic vagotomy on the expression of cfos protein in the PVN and SON following either IP or N injections of 40 pg LPS. Asterisk (*) and double asterisks (**) indicate significantly different from sham-operated animals given IP LPS. Triple asterisks (* * *) indicate significantly different from sham-operated animals given N LPS.

VAGOT+FS

SHAM+RES

VAGOT+RES

GROUPS

FIG. 6. Effects of subdiaphragmatic vagotomy on c-fos expression in the PVN and SON induced by 60 min of either intermittent foot shock (+FS) or restraint (+RES). Asterisk (*) indicates significantly different from sham-operated animals exposed to foot shock.

12

of cimetidine has been shown to block central H2 receptors, yet cimetidine had no effect on c-fos induction produced by LPS. Most likely, LPS stimulates the peripheral release of histamine which contributes to the induction of c-fos expression via Hl histamine receptors. The extent to which this effect is linked with the production of cytokines such as TNF or IL-l remains to be determined. Likewise, the effects of central injections of histamine antagonist on the induction of c-fos protein by LPS has not been tested. Glutamate represents the major excitatory neural transmitter in the central nervous system and, in addition, is a primary mediator of afferent sensory information. MK801, a NMDA glutamate antagonist, has anticonvulsant properties and potent noncompetitive antagonistic actions on depolarization induced by NMDA (20,68). We found that MK801 blocked the expression of c-fos protein induced by both LPS and foot shock. It has been reported that c-fos induction produced by seizures, ischaemia, traumatic brain injury, and long-term potentiation all involve the activation of NMDA receptors (46) and the distribution of c-fos protein after seizures reflects the location of NMDA receptors (30). Systemic administration of MK801 can also protect neurons from NMDA-mediated excitotoxicity (38). Our results further support the fundamental role of excitatory amino acids in the activation of c-fos protein synthesis in the brain. NMDA stimulates corticosterone release, suggesting that CRF neurons either possess NMDA receptors and/or receive neural inputs from cells that are activated via NMDA receptors (49,54). However, because lower doses (1.0 or 3.0 mg/kg ) of MK801 had only a modest effect on foot shock-induced c-fos expression (unpublished observation) and a high dose of MK801 (5 mgkg) was required to inhibit c-fos induction, other mechanisms (nonNMDA receptors) may also be involved. Similar to previous reports (lo), we also found that MK801 treatment alone induced numerous c-fos-positive neurons in specific brain areas, such as the thalamus, as well as some cells located in the PVN. Because systemic injections of MK801 are reported to stimulate CRF and corticosterone secretion (58) drug treatment alone may constitute a moderately stressful stimulus. Although the present results show that glutamate is a significant mediator of the functional effects of both endotoxin and stress, the specific site(s) of action is not known. However, it is likely that MK801 may be acting at multiple sites including primary sensory afferents, central chemoreceptors, as well as at the level of individual neurons that express c-fos protein. The sequence of events leading to the induction of c-fos protein in the brain includes the depolarization of neurons, release and binding of neurotransmitter to postsynaptic receptors, activation of second messengers systems, and transcription/translation of c-fos (6). Although the exact central neurotransmitter systems represented by the c-fos-positive neurons identified in the present study remains unknown, studies have suggested that vasopressin, oxytocin, CRF, ACTH, and enkephalin neurons may be involved. CRF neurons, fibers, and receptors have been identified in the PVN (7,50,53), and the gastric and colonic motor responses to restraint stress are mediated via CRF (29). Thus, a widely distributed and integrated central CRF system may represent a central mediator of neuroendocrine, autonomic, and behavioral responses to stressful stimuli (5,22). In addition, ACTH and enkephalin immunoreactive fibers/terminals have been shown to be colocalized with c-fos-positive neurons in the PVN following mustard oil application (41). We are currently conducting similar colocalization experiments and preliminary results (Jackson and Nance, unpublished data) indicate many of the cells in the PVN and SON that express c-fos protein following LPS injections are also positive for vasopressin, oxytocin, and

WAN ET AL.

NADPH-diaphorase, a histochemical marker for nitric oxide-producing neurons (4). We have demonstrated that the functional effects of bilateral abdominal vagotomy on c-fos protein was dependent upon the specific stimulus utilized to activate c-fos expression in the brain. Whereas vagotomy completely blocked the ability of IP injections of LPS to induce c-fos expression in the PVN and SON of the rat brain, it had minimal effects on the induction of c-fos protein produced by IV injections of LPS and actually potentiated the activational effects of foot shock on hypothalamic c-fos protein. This latter observation may represent a functional neuroanatomical correlate of the reported role of vagal afferents in nociception. Stimulation of subdiaphragramatic vagal afferents has been shown to produce an intensity-dependent increase in analgesia in response to noxious stimulation (55) and consistent with this, vagotomy produces hyperalgesia (42). Because the number and staining intensity of c-fos-positive cells in the brain induced by foot shock is a positive function of stimulus intensity (Wetmore and Nance, unpublished data), the observed increase in the number of c-fos-positive neurons in the hypothalamus of vagotomized animals corraborate these observations. Finally, the fact that stress, seizures, and LPS injections are reported to induce IL-1 expression in the brain (2,14,27,28) suggests that this cytokine may represent another common central mediator. However, our results indicate that if the central production of IL-1 is related to the activational effects of stress on the hypothalamus and the induction of c-fos protein in the brain, it is probably independent of prostaglandin synthesis. Neurons in the PVN project to the pituitary, median eminence, and brain stem nuclei, such as the nucleus of the tractus solitarius (NTS/A2 region), the dorsal motor nucleus (DMN) of the vagus, as well as the spinal cord. Thus, the PVN represent a pivotal region for integration and control of the endocrine system and the parasympathetic and sympathetic nervous system (35,52,53). Connections between the PVN and the preganglionic vagal efferents may be involved in controlling gastric secretion, and the PVN may exert both inhibitory and excitatory effects on gastric activity via either oxytocinergic or thyrotropin releasing hormone neurons that project directly to the dorsal motor nucleus of the vagus (45). Importantly, bilateral cervical or abdominal vagotomy have been shown to completely abolish the responses of PVN neurosecretory cells to gastric distension and electrical stimulation (36,61). Systemic injections of the gut peptide cholecystokinin (CCK) decreases gastric motility, stimulates pituitary release of oxytocin, and activates c-fos expression in neurons located in the PVN, SON, and dorsal vagal complex, as well as the central amygdaloid nucleus (39). This suggests that c-fos expression may reflect oxytocin release from these neurons and implicates oxytocin as an important neural modulator of vagal efferent traffic to the gastrointestinal tract. The PVN receives afferent input directly from abdominal viscera via the nucleus tractus solitarius (NTS) and/or via a polysynaptic route involving the Al catecholaminergic neurons in the ventrolateral medulla (36,52,53). Our demonstration that LPS induces c-fos expression in the PVN and the AUNTS (A2) regions of the brain stem that is comparable to the pattern observed following systemic CCK, combined with the fact that vagotomy blocks c-fos induction, suggests that c-fos expression in the brain represents the functional pathway through which LPS alters gastrointestinal function (60). In summary, the present results show that there is a common neural substrate that is activated by a variety of stimuli that generate potent autonomic and neuroendocrine responses. However, it is clear that there are different neural pathways through which this common central circuit can be activated. The stimulation of

REGULATION

13

OF BRAIN c-fos

vagal visceral afferents produced by IP LPS activates this common hypothalamic-brainstem circuit at the level of the dorsal vagal complex, whereas Iv injections of LPS probably access this same central circuit via circumventricular organs in the hypothalamus and/or brain stem. Finally, the activation of this common central circuit by stress is potentially mediated via direct and indirect extrahypothalamic inputs from limbic structures such as the lateral septal area and amygdala onto specific hypo-

thalamic and brain stem nuclei. Clarification of the functional role of various cytokines, whether derived from activated macrophage or induced directly in the central nervous system, awaits further examination. ACKNOWLEDGEMENTS

This work was supported by grants from the MRC of Canada 11589; MT11659) and the NIMH (ROlMH43778-04A2).

(MT-

REFERENCES 1. Abe, H.; Rusak, B.; Robertson, H. A. NMDA and non-NMDA receptor antagonists inhibit photic induction of fos protein in the hamster suprachiasmatic nucleus. Brain Res. Bull. 28(5):831-835; 1992. 2. Ban, E.; Haour, F.; Lenstra, R. Brain interleukin 1 gene expression induced by peripheral lipopolysaccharide administration. Cytokine 4:48-54; 1992. 3. Beutler, B.; Krochin, N.; Milsark, I. W.; Luedke, C.; Cerami, A. Control of cachectin (tumor necrosis factor) synthesis: Mechanisms of endotoxin resistence. Science 232:977-980; 1986. 4. Bredt, D. S.; Glatt, C. E.; Hwang, P. M.; Fotuhi, M.; Dawson, T. M.; Synder, S. H. Nitric oxide synthase protein and mRNA are discretely localized in neuronal populations of the mammalian CNS together with NADPH diaphorase. Neuron 7:615-642; 1991. 5. Brown, M. R.; Fisher, L. A. Corticotrophin-releasing factor: Effect on the autonomic system and visceral systems. Fed. Proc. Fed. Am. Sot. Exp. Biol. 44:243-248; 1985. 6. D’Costa, A.; Breese, C. R.; Boyd, R. L.; Booze, R. M.; Sonntag, W. E. Attenuation of fos-like immunoreactivity induced by a single electroconvulsive shock in brains of aging mice. Brain Res. 567:204-211; 1991. 7. De Souza, E. B.; Insel, T. R.; Perrin, M. H.; Rivier, J.; Vale, W. W.; Kuhar, M. J. Corticotrophin-releasing factor receptors are widely distributed within the rat central nervous system: An autoradiographic study. J. Neurosci. 5:3189-3203; 1985. 8. Dinarello, C. A.; Wolff, S. M. Molecular basis of fever in humans. Am. J. Med. 72799-819; 1982. 9. Dragunow, M.; Faull, R. The use of c-fos as a metabolic marker in neuronal pathway tracing. J. Neurosci. Methods 29:261-265; 1989. 10. Dragunow, M.; Faull, R. L. M. MK801 induced c-fos protein in thalamic and neocortical neurons of rat brain. Neurosci. Lett. 111:39-45; 1990. 11 Dragunow, M.; Faull, R. L. M.; Jansen, K. L. R. MK801, an antagonist of NMDA receptors, inhibits injury-induced c-fos protein accumulation in rat brain. Neurosci. Lett. 109:128-133; 1990. 12 Dragunow, M.; Robertson, G. S.; Faull, R. L. M.; Robertson, H. A.; Jansen, K. Dz dopamine receptor antagonists induce fos and related proteins in rat striatal neurons. Neuroscience 37:287-294; 1990. 13 Endres, S.; Cannon, J. G.; Ghorbani, R.; Dempsey, R. A.; Sisson, S. D.; Lonnemann, G.; Van Der Meer, J. W. M.; Wolff, S. M.; Dinarello, C. A. In vitro production of IL-1 beta, IL-1 alpha, TNF and IL-2 in healthy subjects: Distribution, effect of cycooxygenase inhibition and evidence of independent gene regulation. Eur. J. Immunol. 19:2327-2333; 1989. 14. Fontana, A.; Weber, E.; Dayer, J.-M. Synthesis of interleukin 1/ endogeneous pyrogen in the brain of endotoxin-treated mice: A step in fever induction? J. Immunol. 133:1696-1698; 1984. 15. Fritschy, J.-M.; Frondoza, C. G.; Grzanna, R. Differential effects of reserpine on brainstem catecholaminergic neurons revealed by fos protein immunohistochemistry. Brain Res. 562:48-56; 1991. 16. Haliday, E. M.; Ramesha, C. S.; Ringold, G. TNF induces c-fos via a novel pathway requiring conversion of arachidonic acid to a lipoxygenase metabolite. EMBO J. 10(1):109-115; 1991. 17. Herrera, D. G.; Robertson, H. A. N-Methyl-o-aspartate receptors mediate activation of the c-fos proto-oncogene in a model of brain injury. Neuroscience 35(2):273-281; 1990. 18. Hough, L. B.; Glick, S. D.; Su, K. Cimetidine penetrates brain and inhibits nonopiate foot shock-induced analgesia. Pharmacol. B&hem. Behav. 24(5):1257-1261; 1986.

19. Hunt, S. P.; Pini, A.; Evans, G. Induction of c-fos-like protein in spinal tort neurons following sensory stimulation. Nature 328:632634; 1987. 20. Kemp, J. A.; Foster, A. C.; Wong, E. H. Noncompetitive antagonists of excitatory amino acid receptors. Trends Neurosci. 10:294-298; 1987. 21. Kiss, J. Z. Dynamism of chemoarchitecture in the hypothalamic paraventricular nucleus. Brain Res. Bull. 20(6):699-708; 1988. 22. Koob, G. F.; Bloom, F. E. Corticotrophin-releasing factor and behaviour. Fed. Proc. Fed. Am. Sot. Exp. Biol. 44259-263; 1985. 23. Kunkel, S. L.; Spengler, M.; May, M. A.; Spengler, R.; Larrick, J.; Remick, D. Prostaglandin E2 regulates mac&phage-derived tumour necrosis factor gene exoression. J. Biol. Chem. 263:5380-5384: 1988. 24. Levine, J. D.; Fields, H. L.; Basbaum, A. I. Peptides and the primary afferent nociceptor. J. Neurosci. 13:2273-2286, 1993. 25. Masana, M. I.; Heyes, M. P.; Mefford, I. N. Indomethacin prevents increased catecholamine turnover in rat brain following systemic endotoxin challenge. Prog. Neuropsychopharmacol. Biol. Psychiatry 14:609-621; 1990. 26. Mehmet, H.; Millar, J. B. A.; Lehmann, W.; Higgins, T.; Rozengurt, E. Bombesin stimulation of c-fos expression and metagenesis in Swiss 3T3 cells: The role of prostaglandin EZmediated cyclic AMP accumulation. Exp. Cell Res. 190:265-270; 1990. 27. Minami, M.; Kuraishi, Y.; Yamaguchi, T.; Nakai, S.; Hirai, Y.; Satoh, M. Convulsants induce interleukin-8 messenger RNA in rat brain. Biochem. Biophys. Res. Commun. 171:832-837; 1990. 28. Minami, M.; Kuraishi, Y.; Yamaguchi, T.; Nakai, S.; Hirai, Y.; Satoh, M. Immoibilization stresws induces interleukin-l/l mRNA in the rat hypothalamus. Neurosci. Lett. 123:254-256; 1991. 29. Monnikes, H.; Schmidt, B. G.; Raybould, H. E.; Tache, Y. CRF in the paraventricular nucleus mediates gastric and colonic motor response to restraint stress. Am. J. Physiol. 262:G137-G143; 1992. 30. Morgan, J. I.; Cohen, D. R.; Hempstead, J. L.; Curran. T. Mapping patterns of c-fos expression in the central nervous system after seizure. Science 237:192-197; 1987. 31 Morgan, J. I.; Curran, T. Stimulus-transcription coupling in neurons: Role of cellular immediate-early genes. Trends Neurosci. 12:459462; 1989. 32. Morgan, J. I.; Curran, T. Proto-oncogene transcription factors and epilepsy. Trends Pharmacol. 12343-349, 1991. 33. Morrison, D. C.; Ryan, J. L. Endotoxin and disease mechanisms. Annu. Rev. Med. 38:417-432; 1987. 34. Muller, R.; Curran, T.; Mullar, D.; Guilbert, L. Induction of c-fos during myelomonocytic differentiation and macrophage proliferation. Nature 314:546-548; 1985. 35. Nilaver, G.; Zimmerman, E. A.; Wilkins, J.; Michaels, J.; Hoffman, D.; Silverman, A. J. Magnocellular hypothalamic projections to the lower brain stem and spinal tort of the rat. Neuroendocrinology 30:150-158; 1980. 36. Nosaka, S. Solitary nucleus neurons transmitting vagal visceral input to the forebrain via a direct pathway in rats. Exp. Neurol. 85:493505; 1984. 37. Okamoto, H.; Nakano, K. Regulation of interleukin-1 synthesis by histamine produced by mouse peritoneal macrophages per se. Immunology 69:162-165; 1990. 38. Olney, J. W.; Price, M. T.; Salles, K. S.; Labruyere, J.; Frierdich, G. MK801 powerfully protects against N-Methyl-D-Aspartate neurotoxicity. Eur. 3. Pharmacol. 41:357-361; 1987.

14

39. Olson, B. R.; Hoffman, G. E.; Sved, A. F.; Stricker, E. M.; Verbalis, J. G. Cholecystokinin induces c-fos expression in hypothalamic oxytocinergic neurons projecting to the dorsal vagal complex. Brain Res. 569:U8-248; 1992. 40. Panettieri, R. A.; Yadvish, P. A.; Kelly, A. M.; Rubinsterin, N. A.; Kotlikoff, M. 1. Histamine stimulates proliferation of airway smooth muscle and induces c-fo.s expression. Am. J. Physiol. 259:L365371; 1990. 41. Pretel, S.; Piekut, D. T. ACI’H and enkephalin axonal input to para%ricular neurcns containing c-for-like immunoreactivity. Synapse S:lOO-106; 1991. 42. Randich, A.; Robertson, 3. D.; Willingham, T. The use of specific opioid agonist and antagonists to delineate the vagally mediated antinociceptive and cardiovascular effects of intravenous morphine. Brain Res. 603:186-200; 1993. 43 Revhaug, A.; Michie, H. R.; Manson, J. Mck.; Waters, J. M.; Dinarello, C. A.; Wolff, S. M.; Wilmore, D. W. Inhibition of cyclooxygenase attenuates the metabolic response to endotoxin in humans. Arch. Surg. 123:162-170; 1988. 44. Robiion, I. C. A. F. The magnocellular and parvocellular OT and AVP systems. In: Lightman, S. L. et al., eds. Neuroendocrinology. St. Louis, MO: Boston Blackwell Science Publishing; 1986:154-176. 45 Rogers, R. C.; Hermann, G. E. Oxytocin, oxytocin antagonist, TRH, and hypothalamic paraventricular nucleus stimulation effects on gastric motility. Peptides 8(3):505-513; 1987. 46. Rothman, S. M.; Olney, J. W. Excitotoxicity and the NMDA receptor. Trends Neurosci. 10:299-302, 1987. 47. Sager, S. M.; Price, K. J. Systemic endotoxin induced c-fos gene expression in multiple brain regions. Sot. Neurosci. Abstr. 17: 1991. 48. Sawchenko, P. E. Evidence for differential regulation of corticotrophin-releasing factor and vasopressin immunoreactivities in parvocellular nemosecretory and autonomic-related projections of the paraventricular nucleus. Brain Res. 437:253-263; 1987. 49. Sharp, J. W.; Sagar, S, M.; Hisanaga, K.; Jasper, P.; Sharp, F. R. The NMDA receptor mediates cortical induction of c-fos and C-&Xrelated antigens following cortical injury. Exp. Neurol. 109:323332; 1990. 50. Silverman, A. J.; Hou-Yu, A.; Chen, W. P. Corticotrophin-releasing factor synapses within the paraventricular nucleus of the hypothalamus. Neuroendocrinology 49291-299; 1989. 51. Suzuki, S.; Nakano, K. Suppression of endotoxin-induced corticosterone secretion in rats by Hl-~tihistamine. Am. J. Physiol. 248:E26-E30; 1985. 52. Swanson, L. W.; Kuypers, H. G. J. M. The paraventricular nucleus of the hypothalamus: Cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labelling methods. J. Comp. Neurol. 194.555-569; 1980. 53. Swanson, L. W.; Sawchenko, P. E. Hypothalamic integration: Organization of the paravent~~lar and supraoptic nuclei. Annu. Rev. Neurosci. 6:269-324; 1983. 54. Szekely, A. M.; Barbaccia, M. L.; Alho, H.; Costa, E. In primary cultures of cerebellar granule cells the activation of N-methyl-oAspartate sensitive glutamate receptors induces c-fos mRNA expression. Mol. Pharmacol. 35:401-408; 1989. 55. Thurston, C. L.; Randich, A. Electrical stimulation of the subdiaphragmatic vagus in rats: I~ibition of heat-evoked responses of

WAN ET AL.

56.

57.

58

59.

60.

61.

62.

63. 64.

65.

66. 67. 68.

69.

70. 71.

spinal dorsal horn neurons and central substrates mediating inhibition of nociceptive tail flick reflex. Pain 51:349-365, 1992. Tkacs, N. C.; Strack, A. M. Fos immunoreactivity is induced in rat spinal cord autonomic areas following li~~lysacc~ide (LPS) injection. Sot. Neurosci. Abstr. 18:1179; 1992. Torres, G.; Rivier, C. Cocaine-induced expression of striatal c-fos in the rat is inhibited bv NMDA receutor antagonists. Brain Res. Bull. 30:173-176; 1993: Torres, G.; Rivier, C.; Weiss, F. Effects of cocaine on ACTH secretion, strital c-fos expression, and locomotor sensitization in awake and anaesthetized rats. Sot. Neurosci. Abstr. 18:670, 1992. Tsuji, K.; Uehara, A.; Okumura, T.; Taniguchi, Y.; Kitamori, S.; Takasugi, Y.; Namiki, M. The gastric antisecretory action of lipopolysaccharide is blocked by indomethacin. Eur. J. Pharmacol. 210:213-215; 1992. Uehara, A.; Okumura, T.; Okumura, K.; Takasugi, Y.; Namiki, M. Lipopolysaccharide-induced inhibition of gastric acid and pepsin secretion in rats. Eur. J. Pharmacol. 181:141- 1990. Ueta, Y.; Kannan, H.; Yamashita, H. Gastric afferents to the paraventricular nucleus in the rat. Exp. Brain Res. 84:487-494; 1991. Vaillier, D.; Daculsi, R.; Gualde, N. Effects of lipopolysaccharide on interleukin-2-induced cytotoxic activity of murine splenocyte cultures: Role of prostaglandin E2 and interferons. Cancer Immunol. Immunother. 35395-400, 1992. Vannier, E.; Miller, L. C.; Dinarello, C. A. Histamine suppresses gene expression and synthesis of tumour necrosis factor alpha via histamine HZ receptors. J. Exp. Med. 174:281-284; 1991. Wan, W.; Janz, L.; Vriend, C. Y.; Sorensen, C. M.; Greenberg, A. H.; Nance, D. M. Differential induction of c-fos immunoreactivity in hypothalamus and brain stem nuclei following central and peripheral administration of endotoxin. Brain Res. Bull. 32:581587; 1993. Wan, W.; Vriend, C. Y .; Wetmore, L.; Gartner, J. G.; Greenberg, A. H.; Nance, D. M. The effects of stress on splenic immune function are mediated by the splenic nerve. Brain Res. Bull. 3O:IOl-105; 1993. Wetmore, L.; Wan, W.; Nance, D. M. Stress induced c-fos protein in the brain: Temporal-spatial patterning. Sot. Neurosci. Abstr. 18: 1992. Winer, G. Statistical principles in experimental design. New York M~raw-Hilt 1971:267-272. Wong, E. H.; Kemp, J. A.; Priestly, T.; Knight, A. R.; Woodruff, G. N.; Iversen, L. L. The anticonvulsant MK801 is a potent NMethyl-D-Aspartate antagonist. Proc. Natl. Acad. Sci. USA 83: 7104-7108; 1986. Yamashita, H.; Inenaga, K.; Koizumi, K. Possible projections from regions of paraventricular and supraoptic nuclei to the spinal cord: Elec~ophysiological studies. Brain Res. 296373-378; 1984. Zuckerman, S. H.; Evans, G. F.; Butler, L. D. Endotoxin tolerance: Independent regulation of interleukin-1 and tumor necrosis factor expression. Infect. Immun. .59:2774-2780; 1991. Zuckerman, S. H.; Shellhaas, J.; Butler, L. D. Differential regulation of lipopolysaccharide-induced interleukin 1 and tumor necrosis factor synthesis: Effects of endogenous and exogenous glucocorticoids and the rote of the pituitary-adrenal axis. Eur. J. Immunol. 19:301305; 1989.