PII S0361-9230(01)00445-2

Pro-inflammatory and anti-inflammatory cytokinemRNA induction in the periphery and brain following

intraperitoneal administration of bacteriallipopolysaccharide

Nicolas P. Turrin,1 Dave Gayle,1 Sergey E. Ilyin,1 Mark C. Flynn,1 Wolfgang Langhans,2 Gary J. Schwartz3 andCarlos R. Plata-Salaman1*

1Division of Molecular Biology, School of Life and Health Sciences, and Neuroscience Program, University ofDelaware, Newark, DE, USA; 2Institute of Animal Sciences, Swiss Federal Institute of Technology, Zurich,

Switzerland; and 3Bourne Laboratory, Department of Psychiatry, WMC Cornell University,White Plains, NY, USA

[Received 11 September 2000; Revised 28 December 2000; Accepted 14 January 2001]

ABSTRACT: Gram-negative bacteria-derived lipopolysaccha-ride (LPS or endotoxin) is known to play an important role inimmune and neurological manifestations during bacterial infec-tions. LPS exerts its effects through cytokines, and peripheralor brain administration of LPS activates cytokine production inthe brain. In this study, we investigated cytokine and neuropep-tide mRNA profiles in specific brain regions and peripheralorgans, as well as serum tumor necrosis factor (TNF)-a proteinlevels, in response to the intraperitoneal administration of LPS.For the first time, the simultaneous analysis of interleukin(IL)-1b system components (ligand, signaling receptor, recep-tor accessory proteins, receptor antagonist), TNF-a, transform-ing growth factor (TGF)-b1, glycoprotein 130 (IL-6 receptor sig-nal transducer), OB protein (leptin) receptor, neuropeptide Y,and pro-opiomelanocortin (opioid peptide precursor) mRNAswas done in samples from specific brain regions in response toperipherally administered LPS. The same brain region/organsample was assayed for all cytokine mRNA components. Pe-ripherally administered LPS up-regulated pro-inflammatory cy-tokine (IL-1b and/or TNF-a) mRNAs within the cerebral cortex,cerebellum, hippocampus, spleen, liver, and adipose tissue.LPS also increased plasma levels of TNF-a protein. LPS did notup-regulate inhibitory (anti-inflammatory) cytokine (IL-1 recep-tor antagonist and TGF-b1) mRNAs in most brain regions (ex-cept for IL-1 receptor antagonist in the cerebral cortex and forTGF-b1 in the hippocampus), while they were increased in theliver, and IL-1 receptor antagonist was up-regulated in thespleen and adipose tissue. Overall, peripherally administeredLPS modulated the levels of IL-1b system components withinthe brain and periphery, but did not affect the neuropeptide-related components studied. The data suggest specificity oftranscriptional changes induced by LPS and that cytokine com-ponent up-regulation in specific brain regions is relevant to theneurological and neuropsychiatric manifestations associatedwith peripheral LPS challenge. © 2001 Elsevier Science Inc.

KEY WORDS: Endotoxin, Interleukin, Tumor necrosis factor,Growth factor, Neuroimmunology, Neuropeptides, Hypothala-mus, Cortex, Cerebellum, Hippocampus, Liver, Spleen, Adiposetissue, Rat.

INTRODUCTION

Gram-negative bacterial infections are associated with immuno-logical and neurological manifestations including anorexia, fever,sleep changes, learning disabilities, neurological dysfunction, andinjury. Bacterial lipopolysaccharide (LPS or endotoxin) is respon-sible for multiple pathophysiological changes associated withGram-negative infections [24]. LPS acts through pro-inflammatorycytokines including interleukin (IL)-1b and tumor necrosis factor(TNF)-a [7,20,22,32,34,42,45]. Moreover, the exogenous admin-istration of a competitive endogenous inhibitor, IL-1 receptorantagonist (IL-1Ra), blocks or ameliorates the central nervoussystem (CNS) effects induced by (central) LPS and IL-1b (e.g.,[2,4,19,48]).

Models of peripheral immune challenge or peripheral inflam-mation have shown cytokine induction within the brain [8–10,23,28,31,36,37,44,56,57,63,65]. The exact mechanism by which braincytokines are regulated during peripheral infections is still underdebate; possible models include: (i) intermediate soluble factors,such as prostaglandins or cytokines themselves, produced andreleased within the CNS (e.g., by microglia), and/or from endo-thelial cells of the cerebrovasculature and/or from circumventricu-lar organs (CVOs) in response to LPS [3,21,31,40]; (ii) transport ofcytokines across the blood—brain barrier and/or CVOs, or directactivation of CVO neurons [1,40]; and (iii) neural afferent signal-ing inducing CNS cytokine production [16,38,39]. However, con-troversy exists with the model of vagal [16,38] or sympatheticsignaling because this model has not been supported by experi-

* Address for correspondence: Carlos R. Plata-Salama´n, M.D., D.Sc., Central Nervous System Research, The R. W. Johnson Pharmaceutical ResearchInstitute, Welsh and McKean Roads, Spring House, PA 19477-0776, USA. Fax:11-(215)-628-3297; E-mail: CPlataSa@prius.jnj.com

Brain Research Bulletin, Vol. 54, No. 4, pp. 443–453, 2001Copyright © 2001 Elsevier Science Inc.Printed in the USA. All rights reserved

0361-9230/01/$–see front matter

443

ments using selective subdiaphragmatic deafferentation or celiacsuperior mesenteric ganglionectomy [55,58], suggesting that hu-moral mediation may play a critical role. Brain administration ofLPS induces IL-1b and other cytokines in specific brain regions[5,13,15,51]. Peripheral injections of LPS are also able to inducecentral cytokine production (and central cytokines mediate theCNS effects induced by peripheral LPS) [2,19,23,28,29,31,36,44,56,57,63]. Other studies have also reported the modulation ofcytokine components by LPS [11,12,63]. It should be noted, how-ever, that several of these studies used extremely high doses ofLPS (in some cases mg/kg) (e.g., [11,29,32]). Moreover, the si-multaneous analysis of multiple cytokine systems and their com-ponents (ligands, receptors, signal transducing proteins, inhibitorymechanisms) within specific brain regions in the same animalfollowing a peripheral LPS challenge has not been characterized.This characterization of brain cytokine component profiles follow-ing LPS peripheral administration is pivotal to understand periph-eral-to-central immune signaling as well as the neurological andneuropsychiatric manifestations associated with bacterial productand cytokine peripheral challenges [34,41,43,48].

In the present study, the cytokine mRNA profiles (as evidencefor local cytokine production within the brain) in response to theintraperitoneal administration of LPS were investigated. Two pro-inflammatory cytokines (IL-1b and TNF-a), two anti-inflamma-tory or inhibitory cytokines (IL-1Ra and transforming growthfactor [TGF]-b1), IL-1b system-associated components (IL-1 re-ceptor type I or signaling receptor, and IL-1 receptor accessoryprotein [IL-1R AcP] I and II—which increase the binding affinityof IL-1b for IL-1RI when the two proteins are co-expressed [17];IL-1R AcP expression also correlates with IL-1 responsiveness)[64], glycoprotein 130 (gp 130, a common signal transduceramong receptors for members of the IL-6 receptor subfamily [46],and the signaling OB receptor (OB-R or leptin receptor, whichshows homology with gp 130 and responds to leptin, an adipositycytokine signal that inhibits feeding and that can be modulated bycytokines [18]) mRNA levels were examined in specific brainregions (cerebral cortex, cerebellum, hippocampus, and hypothal-amus) and peripheral organs (liver, spleen, and adipose tissue).Hence, the simultaneous analysis of IL-1b system components,TNF-a, TGF-b1, and gp 130 mRNAs can provide information onmultiple cytokine components and on potential cytokine-cytokineinteractions. Cytokine-associated mRNAs were examined in thesame brain region and peripheral organ samples.

Cytokine-neuropeptide Y and cytokine-endogenous opioid in-teractions may play a role in cytokine-associated CNS activities[30,47,60,61]. Thus, mRNA levels for neuropeptide Y (NPY) andpro-opiomelanocortin (POMC, opioid peptide precursor) in re-sponse to intraperitoneally (i.p.) administered LPS were also stud-ied. Because LPS induces anorexia, analysis of POMC gene ex-pression also could provide insights into potential changes ofmelanocortin-related mechanisms.

MATERIALS AND METHODS

Subjects and Maintenance

Male Wistar (VAF) rats weighing between 250 and 275 g at thebeginning of the experiments were used. Rats were randomlyassigned into groups, housed individually, and maintainedadlibitum on powdered rat food (Labdiet, PMI Feeds, Inc., St. Louis,MO, USA) and tap water [53]. Lights were on from 0600 to 1800 hand room temperature was kept at 236 1°C. All rats were handleddaily.

Bacterial Lipopolysaccharide Administration

The rats were injected i.p. using either vehicle (sterile physio-logical saline) or 100mg/kg of bacterial lipopolysaccharide (Esch-erichia coli serotype O111:B4; Sigma, St. Louis, MO, USA; Lot78H4086) in sterile physiological saline. The same lot and stocksolutions were used for all experiments. The concentration of LPSselected was based on previous studies that used the acute i.p.administration [33,35,54]. The time selected for tissue sampling (5h after the acute i.p. administration) was based on previous studieson the cytokine and behavioral profiles exhibited by rats receivingLPS [33,35,51].

Dissection of Brain Regions and Peripheral Organs

Five hours following the i.p. injection, rats were decapitatedand their brains were quickly removed (,30 s from the time ofdecapitation). Trunk blood was also taken from the animals. Eachbrain was immediately placed in oxygenated phosphate-bufferedsaline solution at 2–4°C. Brains were rinsed several times andselected regions (cerebellum, parieto-frontal cortex, hippocampus,and hypothalamus) were dissected. Liver, spleen, and epididymaladipose tissue samples were also taken immediately from the rats.Each brain or peripheral sample was immediately homogenized inguanidine thiocyanate/phenol solution (Tri Reagent; MolecularResearch Center, Inc., Cincinnati, OH, USA) using a microtissuegrinder, and frozen at285°C. The samples were number coded forfurther processing and analyses. The complete brain and organdissection, homogenization and storage took less than 6 min.

RNA Isolation and RNase Protection Assay

Total cell RNA was isolated individually from the homoge-nized samples according to our previous studies [25,51]. RNAconcentration was determined by spectrophotometry at an absor-bance of 260 nm. The RNA integrity was assessed using agarosegel electrophoresis and ethidium bromide staining. The level of ratglyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA andb-actin mRNA were RNase protection assayed to confirm consis-tency and that an equal amount of total cell RNA was used for eachassay.

Riboprobes were prepared as previously described [14,15,26,49]. Probe synthesis was conducted with 1 mM each of cytidine59-triphosphate (CTP), adenosine 59-triphosphate, and uridine 59-triphosphate; and 9.38mM of [32P] guanosine 59-triphosphate(GTP) (800 Ci/mmol) and 25mM of unlabeled GTP. [32P]GTP notincorporated into the probe was removed by two ethanol precipi-tations in the presence of 2.5 M ammonium acetate. RNase pro-tection assays were used to detect the IL-1b, IL-1RI, IL-1R AcPs,IL-1Ra, TNF-a, TGF-b1, gp 130, OB-R, POMC, NPY, GAPDH,andb-actin mRNAs.

Hybridization reactions containing 10.0mg of total cell RNAand 2.53 104 cpm each of IL-1b, IL-1RI, IL-1R AcP, IL-1Ra, andTGF-b1 antisense probes, and 1.53 104 cpm of TNF-a antisenseprobe; 10.0mg of total cell RNA and 2.53 104 cpm each of gp130, OB-R, POMC, and NPY antisense probes, or 3.0mg of totalcell RNA and 2.03 105 cpm each of GAPDH andb-actinantisense probes in 30ml of hybridization buffer (80% formamide,0.4 M NaCl, 1 mM EDTA, 40 mM PIPES [pH 6.4]) were heatedto 85°C for 5 min and then incubated at 48°C for 12–18 h. Afterhybridization, 280ml of RNase digestion buffer (50 mM sodiumacetate [pH 4.5], 2 mM EDTA) was added with 30 U of T1 RNase(Sigma) for all assays followed by incubation at 30°C for 60 min.RNase digestion was terminated by the addition of 10ml of 20%sodium dodecyl sulfate and 50mg of proteinase K and incubationfor 30 min at 37°C. RNA was extracted with phenol/chloroform

444 TURRIN ET AL.

and precipitated with 70mg of yeast transfer RNA by the additionof ethanol. RNA was dissolved in loading buffer (80% formamide,2 mM EDTA [pH 7.4] containing 0.05% bromophenol blue and0.05% xylene cyanol), denatured at 85°C for 5 min, and resolvedon 5% acrylamide/8 M urea gels using TBE-buffer (89 mM TrispH 8.0, 89 mM boric acid and 2.7 mM EDTA). Gels wereautoradiographed and results were quantified with an image ana-lyzer (Image Quant, Molecular Dynamics, Sunnyvale, CA, USA).Densitometric values for each mRNA analyzed were converted topercentage values of the total values for a particular mRNA.

In control experiments on hybridization specificity, 10.0mg ofyeast transfer RNA were hybridized and processed as describedabove. No signal corresponding to IL-1b, IL-1RI, IL-1R AcPs,IL-1Ra, TNF-a, TGF-b1, gp 130, OB-R, POMC, or NPY wasdetected.

Riboprobe Templates

Rat IL-1b expression plasmid containing the entire maturepeptide coding sequence of rat IL-1b, with one extra Met codon atthe 59 end cloned into plasmid pET-21d (Novagen, Madison, WI,USA) between EcoRI and NcoI sites was used. Plasmid rat ribo-probe IL-1b (rRIL-1b) was generated by cloning IL-1b containingXbaI-EcoRI fragment of the rat IL-1b expression plasmid intopGEM-2 between XbaI and EcoRI sites. Plasmid rRIL-1b waslinearized with HindIII, and transcribed with SP6 RNA polymer-ase to generate an approximately 570 nucleotides-long antisenseprobe that protects 492 nucleotides in the rat IL-1b mRNA.

Plasmid rat IL-1RI 39 clone containing a 39 fragment (nucleo-tides 473–1826 of accession # M95578) of the rat IL-1RI cDNAcloned into the SmaI site of pBluescript SK (Stratagene) was used.Linearization of this plasmid with HindIII and transcription withT3 RNA polymerase produced an approximately 520 nucleotides-long antisense probe that protects 436 nucleotides in the rat IL-1RImRNA.

Rat IL-1R AcP cDNA (nucleotides 874–1212 of accession #U48592) cloned into EcoRV site of pBluescript SK was used.Linearization of this plasmid by HindIII and transcription by T3RNA polymerase resulted in an approximately 420 nucleotides-long antisense probe that protects 339 nucleotides in the membranebound form of rat IL-1R AcP; additionally, this probe protects ashorter fragment corresponding to the soluble form of IL-1R AcP[17].

Plasmid pBS-RA containing a fragment (nucleotides 3–451 ofaccession # M63101) of the rat IL-1Ra cDNA cloned into the SmaIsite of pBluescript SK was linearized with XmnI, and transcribedwith T3 RNA polymerase to produce an approximately 250 nu-cleotides-long antisense probe that protects 207 nucleotides in therat IL-1Ra mRNA.

Plasmid prTNFtrans(T7) containing the peptide coding portionof rat TNF-a cDNA (nucleotides 1–708 of accession # S40199)was used. Linearization of this plasmid by EcoRI and transcriptionwith SP6 RNA polymerase produced an approximately 790 nucle-otides-long antisense probe which protects 708 nucleotides in therat TNF-a mRNA.

Rat TGF-b1 cDNA (accession # X52498) cloned into pBlue-script II KS vector (Stratagene) was used. Linearization of thisplasmid by BamHI and transcription with T3 RNA polymeraseproduced an approximately 320 nucleotides-long antisense probethat protects 244 nucleotides in the rat TGF-b1 mRNA.

Rat gp 130 cDNA (accession # M92340) cloned into EcoRI siteof pBS vector was used. This cDNA was linearized with BsaI andtranscribed with T3 RNA polymerase to produce an approximately590 nucleotides-long antisense probe that protects 530 nucleotidesin the rat gp 130 mRNA.

Leptin receptor riboprobe template was generated by cloning aHincII-BamHI fragment of rat leptin receptor (Ob-Rb isoform)cDNA into respective sites of pGEM-2. The leptin receptor ribo-probe was linearized using HindIII and SP6 transcription resultedin an approximately 450 nucleotides-long antisense probe thatprotects 415 nucleotides (corresponding to nucleotides 686-1101of accession # U52966) in the rat OB-R mRNA.

Rat POMC cDNA cloned between BamHI and EcoRI sites ofpBluescript KS(1) vector was used. Plasmid was linearized withXmnI and T3 RNA polymerase transcription resulted in an ap-proximately 320 nucleotides-long antisense probe that protects 264nucleotides in the rat POMC mRNA (predominant form).

Plasmid pBLNPY-1 containing a 511 bp insert (accession #M20373) comprising most of the cDNA of rat prepro-neuropeptideY ligated into the EcoRI site of vector Bluescribe M13(2) wasused. Linearization of the plasmid with BbsI and transcription withT3 RNA polymerase produced an approximately 242 nucleotides-long antisense probe that protects 180 nucleotides in the rat NPYmRNA.

Rat antisense template pTRI-GAPDH (Ambion; which con-tains a 316 bp fragment of the rat GAPDH gene) was used togenerate the GAPDH antisense probe. Antisense template pTRI-b-actin-125-Rat (Ambion; which contains a 126 bp cDNA frag-ment of the ratb-actin gene) was used to generate theb-actinantisense probe.

Rat TNF-a Enzyme-Linked Immunoabsorbent Assay

The blood samples were analyzed with a rat ultra-sensitiveTNF-a enzyme-linked immunoabsorbent assay according to themanufacturer’s protocol (Cytoscreen™ Immunoassay kit; Bio-Source International, Inc., Camarillo, CA, USA). The sensitivity ofthe assay is,0.7 pg/ml (range, 2.3–150 pg/ml), and there is nocross-reactivity with other cytokines.

Data Analyses

Results are expressed as mean6 SE unless otherwise stated.Data were analyzed using analysis of variance (ANOVA) followedby post-hoc tests for pairwise comparisons (Student-Newman-Keuls method) if there was a significant main effect. Kruskal-Wallis ANOVA on ranks was applied (followed bypost-hoctests)when data did not pass the normality (Kolmogorov-Smirnov) andequal variance (Levene Median) tests. When appropriate, datawere also analyzed using thet-test or Mann-WhitneyU-test (whendata did not pass normality and equal variance tests).

RESULTS

RNase Protection Assay

The same animals were used to analyze all components studied:IL-1b, IL-1RI, IL-1R AcP I and II, IL-1Ra, TNF-a, TGF-b1, gp130, OB-R, POMC, and NPY mRNA levels. A sample gel of theRNase protection assays is shown in Fig. 1. As shown, samplesfrom both the vehicle (controls)- and LPS-treated animals fromdifferent brain regions (Fig. 1A) and peripheral organs (Fig. 1B)were analyzed concomitantly. Consistency was verified by analyz-ing all samples individually, from which all means6 SE weregenerated. Each individual sample was obtained from a differentrat, and all rats provided all four brain and three peripheral organsamples; the same samples were analyzed with all antisense probesas described in Materials and Methods. GAPDH andb-actinmRNAs were relatively constant between treatments, with a brainregion suggesting consistency of processing and that equalamounts of total RNA were used. In addition, to verify consistencyof the assay, serial dilutions of a test RNA sample were also

BRAIN CYTOKINES AND LPS 445

analyzed (not shown). The specificity of the probes used has beendemonstrated previously by various procedures, including controlexperiments on hybridization specificity in which no signal corre-sponding to IL-1b, IL-1RI, IL-1R AcPs, IL-1Ra, TNF-a, TGF-b1,gp 130, OB-R, POMC, or NPY was detected [13,15,25,27,49–52].

IL-1b mRNA Levels

Intraperitoneally administered LPS significantly induced theexpression of IL-1b in cerebellum (p , 0.001), parieto-frontalcortex (p , 0.001), and hippocampus (p 5 0.003) (Fig. 2). Thesame trend was observed in the hypothalamus. The cerebellum andparieto-frontal cortex showed an induction of similar magnitude(i.e., the difference between control and LPS-treated animals) of14.7 and 16.0 arbitrary units, respectively.

Intraperitoneal LPS significantly induced IL-1b mRNA expres-sion in the peripheral samples compared to vehicle treatment: liver(p , 0.001), spleen (p , 0.001), and adipose tissue (p , 0.001)(Fig. 2). The magnitude of induction was greatest in the spleen(41.0 arbitrary units difference between control and LPS treat-ments). liver (34.2). adipose tissue (9.5).

IL-1Ra mRNA Levels

There was no significant induction of IL-1Ra mRNA in thecerebellum, hippocampus, or hypothalamus in response to i.p. LPS(Fig. 3). Only the parieto-frontal cortex showed a significantLPS-stimulated up-regulation of IL-1Ra mRNA levels (p 5 0.01).

Intraperitoneal LPS also induced a robust up-regulation ofIL-1Ra mRNA in the liver (magnitude of induction relative tocontrols5 38.6 arbitrary units,p , 0.001) and spleen (25.7,p ,0.001), and a milder, but significant increase of 5.7 arbitrary unitsover control in the adipose tissue (p , 0.001).

IL-1RI mRNA Levels

Intraperitoneal LPS significantly induced cerebellar (p 50.001), cortical (p , 0.001), and hippocampal (p 5 0.001), but nothypothalamic, IL-1RI mRNA expression (Fig. 4). The highermagnitudes of induction were exhibited by the cerebellum andparieto-frontal cortex.

Analysis of peripheral expression also revealed a significantLPS-induced up-regulation of IL-1RI mRNA in liver (p , 0.001),spleen (p , 0.001), and adipose tissue (p , 0.001). The livershowed the highest magnitude of induction.

IL-1R AcP I and II mRNA Levels

Intraperitoneal LPS significantly induced brain IL-1R AcP ImRNA expression (Fig. 5) in the cerebellum (p , 0.04) andparieto-frontal cortex (p , 0.04); no changes were observed in thehippocampus or hypothalamus. All three peripheral tissues showedsignificant LPS-induced up-regulation of IL-1R AcP I mRNA (p ,0.001, liver and spleen;p 5 0.002, adipose tissue). The liver andspleen showed similarly high magnitudes of induction while theadipose tissue exhibited a lower, but still robust increase.

Figure 6 shows a similar trend for the IL-1R AcP II mRNAprofile. Intraperitoneal LPS increased IL-1R AcP II mRNA levelsin cerebellum (p , 0.02), parieto-frontal cortex (p , 0.001), liver(p , 0.001), spleen (p , 0.001), and adipose tissue (p , 0.02).The spleen exhibited the highest magnitude of induction relative tothe vehicle-treated group.

TNF-a mRNA Levels

There was a LPS-stimulated increase of TNF-a mRNA expres-sion (Fig. 7) in the parieto-frontal cortex (p , 0.001) and hip-

FIG. 1. (A) Tumor necrosis factor-a (TNF-a), interleukin-1b (IL-1b),interleukin-1 receptor type I (IL-1RI), interleukin-1 receptor accessoryprotein type I (AcPI), interleukin-1 receptor accessory protein type II(AcPII), and transforming growth factor-b1 (TGF-b1) mRNA levels in thecerebellum (lane 1), parieto-frontal cortex (lane 2), hippocampus (lane 3),and hypothalamus (lane 4). Male Wistar rats were treated with the acuteintraperitoneal administration of vehicle (physiological saline; control) orbacterial lipopolysaccharide (LPS, 100mg/kg). (B) Same as (A) for liver,spleen, and adipose tissue. Brain and peripheral tissue samples werecollected 5 h after vehicle or LPS administration. Figure shows thatsamples from each treatment and brain region or organ were processedconcomitantly. Each individual sample was obtained from a different rat,and the same rat provided all four brain regions and three organ samples;the same samples were analyzed with all antisense probes as described inMaterials and Methods.

446 TURRIN ET AL.

pocampus (p , 0.001), but not in the hypothalamus. LPS treatmentalso significantly induced TNF-a mRNA expression in the liver(p , 0.001) and spleen (p , 0.001). Interestingly, the magnitudeof induction in the spleen (52.8 arbitrary units over control levels)was much greater than that exhibited in the liver (12.0 arbitraryunits).

TGF-b1 mRNA Levels

Analysis of TGF-b1 mRNA levels showed a significant LPS-induced expression in the hippocampal (p , 0.04) and liver (p ,0.001) samples (Fig. 8). No significant changes were observed inthe other samples examined.

FIG. 2. Interleukin-1b (IL-1b) mRNA levels in response to the intraperitoneal administra-tion of vehicle (controls, open bars) or lipopolysaccharide, 100mg/kg (closed bars). Values(means6 SE; n 5 8 for each group) were standardized to arbitrary units. Abbreviations:CLL, cerebellum; CTX, parieto-frontal cortex; HIP, hippocampus; HYP, hypothalamus;LIV, liver; SPL, spleen; ADP, adipose tissue. *p , 0.001; ‡p , 0.01 from vehicle.

FIG. 3. IL-1 receptor antagonist (IL-1Ra) mRNA levels in response to the intraperitonealadministration of vehicle (controls, open bars) or lipopolysaccharide, 100mg/kg (closedbars). Values (means6 SE; n 5 8 for each group) were standardized to arbitrary units.Abbreviations: CLL, cerebellum; CTX, parieto-frontal cortex; HIP, hippocampus; HYP,hypothalamus; LIV, liver; SPL, spleen; ADP, adipose tissue. *p , 0.001; †p , 0.05 fromvehicle.

BRAIN CYTOKINES AND LPS 447

Gp 130, NPY, OB-R, and POMC mRNA Levels

There was no significant LPS-induced modulation of hypotha-lamic mRNA levels of gp 130 (23.86 3.9 versus 26.26 3.4(mean6 SD) arbitrary units for vehicle and LPS, respectively,n 5

8 each), NPY (24.96 1.2 vs. 25.16 3.2 arbitrary units for vehicleand LPS, respectively,n 5 8 each), OB-R (24.16 4.4 vs. 25.967.0 arbitrary units for vehicle and LPS, respectively,n 5 8 each),or POMC (24.06 2.5 vs. 26.06 3.2 arbitrary units for vehicle andLPS, respectively,n 5 8 each).

FIG. 4. Interleukin-1 receptor type I (IL-1RI) mRNA levels in response to the intraperito-neal administration of vehicle (controls, open bars) or lipopolysaccharide, 100mg/kg(closed bars). Values (means6 SE; n 5 8 for each group) were standardized to arbitraryunits. Abbreviations: CLL, cerebellum; CTX, parieto-frontal cortex; HIP, hippocampus;HYP, hypothalamus; LIV, liver; SPL, spleen; ADP, adipose tissue. *p , 0.001; ‡p , 0.01from vehicle.

FIG. 5. Interleukin-1 receptor accessory protein type I (IL-1R AcP I) mRNA levels inresponse to the intraperitoneal administration of vehicle (controls, open bars) or lipopoly-saccharide, 100mg/kg (closed bars). Values (means6 SE; n 5 8 for each group) werestandardized to arbitrary units. Abbreviations: CLL, cerebellum; CTX, parieto-frontalcortex; HIP, hippocampus; HYP, hypothalamus; LIV, liver; SPL, spleen; ADP, adiposetissue. *p , 0.001; ‡p , 0.01; †p , 0.05 from vehicle.

448 TURRIN ET AL.

TNF-a Protein Levels

This study also determined TNF-a protein levels in the serum ofvehicle and LPS-treated rats. Animals injected i.p. with LPS showedhigh levels of TNF-a protein (mean5 56 pg/ml) relative to controlanimals (mean5 3 pg/ml) at 5 h post-LPS administration, that is, atthe point of analysis of all brain and peripheral samples.

DISCUSSION

Bacterial LPS up-regulated pro-inflammatory cytokine (IL-1band TNF-a) mRNAs in the cerebral cortex, hippocampus, andcerebellum, as well as in peripheral organs 5 h after i.p. adminis-tration. LPS-treated animals also exhibited high levels of circulat-ing TNF-a protein compared to vehicle-treated controls. Other

FIG. 6. Interleukin-1 receptor accessory protein type II (IL-1R AcP II) mRNA levels inresponse to the intraperitoneal administration of vehicle (controls, open bars) or lipopoly-saccharide, 100mg/kg (closed bars). Values (means6 SE; n 5 8 for each group) werestandardized to arbitrary units. Abbreviations: CLL, cerebellum; CTX, parieto-frontalcortex; HIP, hippocampus; HYP, hypothalamus; LIV, liver; SPL, spleen; ADP, adiposetissue. *p , 0.001; †p , 0.05 from vehicle.

FIG. 7. Tumor necrosis factor (TNF)-a mRNA levels in response to the intraperitonealadministration of vehicle (controls, open bars) or lipopolysaccharide, 100mg/kg (closedbars). Values (means6 SE; n 5 8 for each group) were standardized to arbitrary units.Abbreviations: CLL, cerebellum; CTX, parieto-frontal cortex; HIP, hippocampus; HYP,hypothalamus; LIV, liver; SPL, spleen; ADP, adipose tissue. *p , 0.001 from vehicle.

BRAIN CYTOKINES AND LPS 449

components of the IL-1b system (IL-1RI, IL-1 AcP I and II) werealso up-regulated in brain regions and periphery. The results ex-emplify the ability of peripheral signals of immune activation tocommunicate with the CNS and activate a cytokine responsewithin specific brain regions. The data also show that neither gp130, OB-R, NPY, or POMC mRNA was modulated during theconditions of testing used. This suggests that although the brainregion transcriptional changes observed in this model were multi-ple, they were specific to cytokine components.

Anti-inflammatory cytokine (IL-1Ra and TGF-b1) mRNAswere also up-regulated within the brain, but only in the cerebralcortex for IL-1Ra and hippocampus for TGF-b1, while both wereup-regulated in the liver; and IL-1Ra was also up-regulated in thespleen and adipose tissue. IL-1Ra and TGF-b1 mRNA up-regula-tion observed in brain regions, however, was mild in comparisonto the robust pro-inflammatory cytokine mRNA response. Basedon previous reports, the lack or a mild response of IL-1Ra andTGF-b1 modulation could be associated with an imbalance to-wards pro-inflammatory cytokines. The net outcome of this im-balance could participate in the magnitude of the CNS (neurolog-ical and neuropsychiatric) manifestations, that is, an inadequateup-regulation of anti-inflammatory components will preclude com-pensation through negative feedback mechanisms. This is alsoconsistent with previous studies that focused on the relationshipbetween pro-inflammatory—anti-inflammatory cytokine balancein the brain and neurological manifestations [49,51,62].

Our previous model that used centrally administered LPS (intothe third cerebral ventricle, 500 ng LPS/rat) [13,15,51] show asimilar profile of pro-inflammatory cytokine component mRNAexpression in specific brain regions. The profile was different,however, in relation to magnitude of cytokine mRNA componentup-regulation as well as brain regional over-expression. LPS ad-ministered centrally induces a large up-regulation of cytokinemRNAs in all brain regions examined (e.g., cerebellum, hip-pocampus, hypothalamus) [13,51]. Moreover, the central admin-

istration of LPS also induces up-regulation of IL-1Ra mRNA[13,51]; however, the ratio IL-1b to IL-1Ra was significantly large[13,51] and the net pro-inflammatory cytokine profile correlatedwith the magnitude of anorexia. The larger over-expression ofbrain cytokine components in response to the brain administrationof LPS (relative to the i.p. administration) can be explained by thedirect LPS action within the brain. Thus, even when LPS actsdirectly within the brain (central administration) or when it re-quires a cascade of events (peripheral administration) such asactivation of humoral and/or neural signals, the overall net result isthe same, that is, pro-inflammatory cytokine induction in the brain.The similarities in brain cytokine component mRNA over-expres-sion could also be related to the similar neurological manifesta-tions that occur with either the peripheral or central administrationof bacterial LPS. However, although hypothalamic IL-1b mRNAwas up-regulated by i.p. LPS, the magnitude of change did notreach statistical significance because of the intra-group variability.

The pathway for peripheral LPS-induced modulation of braincytokine components is still unresolved. The data on increasedTNF-a protein levels in the circulation of LPS-treated rats exhib-iting up-regulation of pro-inflammatory cytokine components inthe periphery and brain regions suggest an involvement, at least inpart, of circulating cytokines as mediators in the present model. Infact, circulating TNF-a has been proposed to play the most im-portant role to trigger ACTH secretion in response to the i.p.administration of LPS [29]. Thus, the consistent increase in levelsof pro-inflammatory cytokine mRNA components in peripheralorgans and discrete brain regions associated with increased bloodlevels of TNF-a protein suggests that LPS induces a series ofevents that results in the local cytokine component up-regulation invarious compartments (peripheral organs and brain). This is sup-ported by our data because we used mRNA as an index of localtranscriptional changes of the cytokine component. Thus, the datasuggest that local production of cytokine components in the braincould be involved in the induction and/or maintenance of neuro-

FIG. 8. Transforming growth factor (TGF)-b1 mRNA levels in response to the intraperi-toneal administration of vehicle (controls, open bars) or lipopolysaccharide, 100mg/kg(closed bars). Values (means6 SE; n 5 8 for each group) were standardized to arbitraryunits. Abbreviations: CLL, cerebellum; CTX, parieto-frontal cortex; HIP, hippocampus;HYP, hypothalamus; LIV, liver; SPL, spleen; ADP, adipose tissue. *p , 0.001; †p , 0.05from vehicle.

450 TURRIN ET AL.

logical manifestations observed during peripheral LPS administra-tion.

A peripheral model of disease involving rats bearing prostateadenocarcinoma tumor cells also showed activation of cytokinecomponents in specific brain regions, and this activation wasassociated with the development of early cancer anorexia [50]. Onthe same line, there could be a relationship with the known clinicalmanifestations that occur during Gram-negative infections or dur-ing LPS challenge including in humans [34,41,43]. The similaritiesbetween two distinct pathophysiological models, oncological andLPS challenge suggest that although these two peripheral modelsmay act differently, both are able to trigger brain production ofpro-inflammatory cytokine components which could act as medi-ators of neurological and neuropsychiatric manifestations. Theresults from an acute inflammatory (LPS challenge) and chronic(oncological) model add validity to the importance of brain cyto-kine production in response to pathophysiological processes.

Implications

The ability of peripherally administered bacterial LPS to pro-duce an up-regulation of brain IL-1b system components, as wellas TNF-a, and to a lesser extent TGF-b1 mRNAs, suggests thepresence of mechanisms in the periphery that are able to signal thebrain and modify the cytokine profile within specific CNS regions.Whether the nature of these mechanisms is humoral or neural isstill to be determined; however, the concomitant up-regulation ofpro-inflammatory cytokine component mRNAs (indicators of localtranscriptional modulation) in peripheral organs and discrete brainregions associated with increased circulating TNF-a protein levelssupports a humoral role in the present model. TNF-a has beenshown to induce its own production and that of IL-1b from variousimmune and brain cell types [6,45], and, therefore, TNF-a couldbe a possible soluble factor involved in the induction of cytokines.The same applies to IL-1b, the predominant soluble form of IL-1[6]. The brain cytokine profile activated by peripheral LPS injec-tions could have implications in the neurology associated with thismodel of peripheral challenge, including in humans. That is, theperipheral administration of bacterial LPS to healthy human vol-unteers results in a dose-dependent induction of neurological man-ifestations [41,43]. The possible relationship of brain regionalcytokine component up-regulation and neurobehavioral conse-quences in response to LPS administration has been discussed[15,51]. Overall, cytokine-cytokine interactions with positive(IL-1b7 TNF-a) [59] and negative (IL-1Ra3 IL-1b; TGF-b3IL-1b/TNF-a) [2,6,48] feedback or a balance between pro-inflam-matory and inhibitory cytokines may be pivotal for the modulationof the cascade of events associated with the neurological manifes-tations of peripherally or centrally administered LPS. A balancetowards pro-inflammatory cytokine profile could be reflected in themagnitude of neurological manifestations and complications [51,62]. Our findings in the presentin vivo model are consistent withthis proposal.

ACKNOWLEDGEMENTS

We thank Dr. Ronald P. Hart (Department of Biological Sciences,Rutgers University) for providing the rat IL-1b, IL-1Ra, IL-1RI, and IL-1RAcP cDNAs; Dr. Karl Decker (Biochemisches Institut der Albert LudwigsUniversitat) for providing the rat TNF-a cDNA; Dr. David Danielpour(National Cancer Institute) for providing the rat TGF-b1 cDNA; Dr. GeraldM. Fuller (Department of Cell Biology and Anatomy, University of Ala-bama at Birmingham) for providing the rat glycoprotein 130 cDNA; Dr.Charles I. Rosenblum (Merck Research Laboratories, Rahway, NJ) forproviding the rat leptin receptor (Ob-Rb isoform) cDNA; Dr. Andrea Gore(Center for Neurobiology, The Mount Sinai Medical Center) for providingthe rat pro-opiomelanocortin cDNA; and Dr. Steven L. Sabol (Laboratory

of Biochemical Genetics, National Heart, Lung, and Blood Institute, Na-tional Institutes of Health) for providing the rat NPY cDNA. Researchsupported by funds from the University of Delaware and the NationalInstitutes of Health to C.R.P.-S. and the E.T.H. to W.L.

REFERENCES

1. Banks, W. A.; Kastin, A. J.; Broadwell, R. D. Passage of cytokinesacross the blood-brain barrier. Neuroimmunomodulation 2:241–248;1995.

2. Cartmell, T.; Luheshi, G. N.; Rothwell, N. J. Brain sites of action ofendogenous interleukin-1 in the febrile response to localized inflam-mation in the rat. J. Physiol. (Lond.) 518:585–594; 1999.

3. Day, H. E.; Akil, H. Differential pattern of c-fos mRNA in the rat brainfollowing central and systemic administration of interleukin-1-beta:Implications for mechanism of action. Neuroendocrinology 63:207–218; 1996.

4. del Rey, A.; Monge-Arditi, G.; Besedovsky, H. O. Central and periph-eral mechanisms contribute to the hypoglycemia induced by interleu-kin-1. Ann. N.Y. Acad. Sci. 840:153–161; 1998.

5. De Simoni, M. G.; Del Bo, R.; De Luigi, A.; Simard, S.; Forloni, G.Central endotoxin induces different patterns of interleukin (IL)-1b andIL-6 messenger ribonucleic acid expression and IL-6 secretion in thebrain and periphery. Endocrinology 136:897–902; 1995.

6. Dinarello, C. A. Biological basis for interleukin-1 in disease. Blood87:2095–2147; 1996.

7. Dinarello, C. A. Cytokines as endogenous pyrogens. J. Infect. Dis.179:S294–S304; 1999.

8. Eriksson, C.; Van Dam, A. M.; Lucassen, P. J.; Bol, J. G.; Winblad, B.;Schultzberg, M. Immunohistochemical localization of interleukin-1be-ta, interleukin-1 receptor antagonist and interleukin-1beta convertingenzyme/caspase-1 in the rat brain after peripheral administration ofkainic acid. Neuroscience 93:915–930; 1999.

9. Eriksson, C.; Winblad, B.; Schultzberg, M. Kainic acid induced ex-pression of interleukin-1 receptor antagonist mRNA in the rat brain.Mol. Brain Res. 58:195–208; 1998.

10. Gabellec, M. M.; Griffais, R.; Fillion, G.; Haour, F. Expression ofinterleukin 1 alpha, interleukin 1 beta and interleukin 1 receptorantagonist mRNA in mouse brain: Regulation by bacterial lipopoly-saccharide (LPS) treatment. Brain Res. Mol. Brain Res. 31:122–130;1995.

11. Gabellec, M. M.; Griffais, R.; Fillion, G.; Haour, F. Interleukin-1receptors type I and type II in the mouse brain: Kinetics of mRNAexpressions after peripheral administration of bacterial lipopolysac-charide. J. Neuroimmunol. 66:65–70; 1996.

12. Gabellec, M. M.; Jafarian-Tehrani, M.; Griffais, R.; Haour, F. Inter-leukin-1 receptor accessory protein transcripts in the brain and spleen;kinetics after peripheral administration of bacterial lipopolysaccharidein mice. Neuroimmunomodulation 3:304–309; 1996.

13. Gayle, D.; Ilyin, S. E.; Flynn, M. C.; Plata-Salama´n, C. R. Lipopoly-saccharide (LPS)- and muramyl dipeptide (MDP)-induced anorexiaduring refeeding following acute fasting: Characterization of braincytokine and neuropeptide systems mRNAs. Brain Res. 795:77–86;1998.

14. Gayle, D.; Ilyin, S. E.; Plata-Salaman, C. R. Central nervous systemIL-1b system and neuropeptide Y mRNAs during IL-1b-inducedanorexia in rats. Brain Res. Bull. 44:311–317; 1997.

15. Gayle, D.; Ilyin, S. E.; Plata-Salaman, C. R. Feeding status andbacterial LPS-induced cytokine and neuropeptide gene expression inthe hypothalamus. Am. J. Physiol. 277:R1188–R1195; 1999.

16. Goehler, L. E.; Gaykema, R. P.; Nguyen, K. T.; Lee, J. E.; Tilders,F. J.; Maier, S. F.; Watkins, L. R. Interleukin-1b in immune cells of theabdominal vagus nerve: A link between the immune and nervoussystems? J. Neurosci. 19:2799–2806; 1999.

17. Greenfeder, S. A.; Nunes, P.; Kwee, L.; Labow, M.; Chizzonite, R. A.;Ju, G. Molecular cloning and characterization of a second subunit ofthe interleukin 1 receptor complex. J. Biol. Chem. 270:13757–13765;1995.

18. Grunfeld, C.; Zhao, C.; Fuller, J.; Pollock, A.; Moser, A.; Friedman, J.;Feingold, K. R. Endotoxin and cytokines induce expression of leptin,

BRAIN CYTOKINES AND LPS 451

the ob gene product, in hamsters. A role for leptin in the anorexia ofinfection. J. Clin. Invest. 97:2152–2157; 1996.

19. Habu, S.; Watanobe, H.; Yasujima, M.; Suba, T. Different roles ofbrain interleukin 1 in the adrenocorticotropin response to centralversus peripheral administration of lipopolysaccharide in the rat. Cy-tokine 10:390–394; 1998.

20. Hagan, P.; Poole, S.; Bristow, A. F. Endotoxin-stimulated productionof rat hypothalamic interleukin-1b in vivo and in vitro, measured byspecific immunoradiometric assay. J. Mol. Endocrinol. 11:31–36;1993.

21. Herkenham, M.; Lee, H. Y.; Baker, R. A. Temporal and spatialpatterns of c-fos mRNA induced by intravenous interleukin-1: Acascade of non-neuronal cellular activation at the blood-brain barrier.J. Comp. Neurol. 400:175–196; 1998.

22. Hernandez, J.; Hidalgo, J. Endotoxin and intracerebroventricular in-jection of IL-1 and IL-6 induce rat brain metallothionein-I and -II.Neurochem. Int. 32:369–373; 1998.

23. Hillhouse, E. W.; Mosley, K. Peripheral endotoxin induces hypotha-lamic immunoreactive interleukin-1 beta in the rat. Br. J. Pharmacol.109:289–290; 1993.

24. Holst, O.; Ulmer, A. J.; Brade, H.; Flad, H.-D.; Rietschel, E. T.Biochemistry and cell biology of bacterial endotoxins. FEMS Immu-nol. Med. Microbiol. 16:83–104; 1996.

25. Ilyin, S. E.; Plata-Salaman, C. R.In vivo regulation of the IL-1bsystem (ligand, receptors I and II, receptor accessory protein andreceptor antagonist) and TNF-a mRNAs in specific brain regions.Biochem. Biophys. Res. Commun. 227:861–867; 1996.

26. Ilyin, S. E.; Plata-Salaman, C. R. HIV-1 gp120 modulates hypotha-lamic cytokine mRNAsin vivo: Implications to cytokine feedbacksystems. Biochem. Biophys. Res. Commun. 231:514–518; 1997.

27. Ilyin, S. E.; Sonti, G.; Gayle, D.; Plata-Salaman, C. R. Regulation ofbrain interleukin-1b (IL-1b) system mRNAs in response to pathophys-iological concentrations of IL-1b in the cerebrospinal fluid. J. Mol.Neurosci. 7:169–181; 1996.

28. Jacobs, R. A.; Satta, M. A.; Dahia, P. L.; Chew, S. L.; Grossman, A. B.Induction of nitric oxide synthase and interleukin-1beta, but not hemeoxygenase, messenger RNA in rat brain following peripheral admin-istration of endotoxin. Brain Res. Mol. Brain Res. 49:238–246; 1997.

29. Kakizaki, Y.; Watanobe, H.; Kohsaka, A.; Suda, T. Temporal profilesof interleukin-1beta, interleukin-6, and tumor necrosis factor-alpha inthe plasma and hypothalamic paraventricular nucleus after intravenousor intraperitoneal administration of lipopolysaccharide in the rat: Es-timation by push-pull perfusion. Endocr. J. 46:487–496; 1999.

30. Kong, L.-Y.; McMillian, M. K.; Hudson, P. M.; Jin, L.; Hong, J.-S.Inhibition of lipopolysaccharide-induced nitric oxide and cytokineproduction by ultralow concentrations of dynorphins in mixed gliacultures. J. Pharmacol. Exp. Ther. 280:61–66; 1997.

31. Konsman, J. P.; Kelley, K.; Dantzer, R. Temporal and spatial relation-ships between lipopolysaccharide-induced expression of Fos, interleu-kin-1beta and inducible nitric oxide synthase in rat brain. Neuro-science 89:535–548; 1999.

32. Kozak, W.; Zheng, H.; Conn, C. A.; Soszynski, D.; van der Ploeg,L. H. T.; Kluger, M. J. Thermal and behavioral effects of lipopoly-saccharide and influenza in interleukin-1b-deficient mice. Am. J.Physiol. 269:R969–R977; 1995.

33. Langhans, W.; Balkowski, G.; Savoldelli, D. Differential feedingresponses to bacterial lipopolysaccharide and muramyl dipeptide.Am. J. Physiol. 261:R659–R664; 1991.

34. Langhans, W. Bacterial products and the control of ingestive behavior:Clinical implications. Nutrition 12:303–315; 1996.

35. Langhans, W.; Savoldelli, D.; Weingarten, S. Comparison of thefeeding responses to bacterial lipopolysaccharide and interleukin-1b.Physiol. Behav. 53:643–649; 1993.

36. Laye, S.; Parnet, P.; Goujon, E.; Dantzer, R. Peripheral administrationof lipopolysaccharide induces the expression of cytokine transcripts inthe brain and pituitary of mice. Mol. Brain Res. 27:157–162; 1994.

37. Luheshi, G. N.; Stefferi, A.; Turnbull, A. V.; Dascombe, M. J.;Brouwer, S.; Hopkins, S. J.; Rothwell, N. J. Febrile response to tissueinflammation involves both peripheral and brain IL-1 and TNF-alphain the rat. Am. J. Physiol. 272:R862–R868; 1997.

38. Maier, S. F.; Goehler, L. E.; Fleshner, M.; Watkins, L. R. The role of

the vagus nerve in cytokine-to-brain communication. Ann. N.Y. Acad.Sci. 840:289–300; 1998.

39. Maier, S. F.; Watkins, L. R. Cytokines for psychologists: Implicationsof bidirectional immune-to-brain communication for understandingbehavior, mood, and cognition. Psychol. Rev. 105:83–107; 1998.

40. Maness, L. M.; Kastin, A. J.; Banks, W. A. Relative contributions ofa CVO and the microvascular bed to delivery of blood-borne IL-1a tothe brain. Am. J. Physiol. 275:R207–R212; 1998.

41. Mullington, J.; Korth, C.; Hermann, D. M.; Orth, A.; Galanos, C.;Holsboer, F.; Pollmacher, T. Dose-dependent effects of endotoxin onhuman sleep. Am. J. Physiol. 278:R947–R955; 2000.

42. Parant, M. A.; Pouillart, P.; Le Contel, C.; Parant, F. J.; Chedid, L. A.;Bahr, G. M. Selective modulation of lipopolysaccharide-induced deathand cytokine production by various muramyl peptides. Infect. Immun.63:110–115; 1995.

43. Pernerstorfer, T.; Schmid, R.; Bieglmayer, C.; Eichler, H. G.; Kapiotis,S.; Jilma, B. Acetaminophen has greater antipyretic efficacy thanaspirin in endotoxemia: A randomized, double-blind, placebo-con-trolled trial. Clin. Pharmacol. Ther. 66:51–57; 1999.

44. Pitossi, F.; del Rey, A.; Kabiersch, A.; Besedovsky, H. Induction ofcytokine transcripts in the central nervous system and pituitary fol-lowing peripheral administration of endotoxin to mice. J. Neurosci.Res. 48:287–298; 1997.

45. Plata-Salaman, C. R. Immunoregulators in the nervous system. Neu-rosci. Biobehav. Rev. 15:185–215; 1991.

46. Plata-Salaman, C. R. Anorexia induced by activators of the signaltransducer gp 130. Neuroreport 7:841–844; 1996.

47. Plata-Salaman, C. R. Anorexia during acute and chronic disease:Relevance of neurotransmitter-peptide-cytokine interactions. Nutrition13:159–160; 1997.

48. Plata-Salaman, C. R. Cytokine-induced anorexia: Behavioral, cellular,and molecular mechanisms. Ann. N.Y. Acad. Sci. 856:160–170; 1998.

49. Plata-Salaman, C. R.; Ilyin, S. E. IL-1b-induced modulation of thehypothalamic IL-1b system, TNF-a and TGF-b1 mRNAs in obese(fa/fa) and lean (Fa/Fa) Zucker rats: Implications to IL-1b feedbacksystems and cytokine-cytokine interactions. J. Neurosci. Res. 49:541–550; 1997.

50. Plata-Salaman, C. R.; Ilyin, S. E.; Gayle, D. Brain cytokine mRNAs inanorectic rats bearing prostate adenocarcinoma tumor cells. Am. J.Physiol. 275:R566–R573; 1998.

51. Plata-Salaman, C. R.; Ilyin, S. E.; Gayle, D.; Flynn, M. C. Gram-negative and Gram-positive bacterial products induce differential cy-tokine profiles in the brain: Analysis using an integrative molecular-behavioral in vivo model. Int. J. Mol. Med. 1:387–397; 1998.

52. Plata-Salaman, C. R.; Ilyin, S. E.; Gayle, D.; Romanovitch, A.; Car-bone, K. M. Persistent Borna disease virus infection of neonatal ratscauses brain regional changes of mRNAs for cytokines, cytokinereceptor components and neuropeptides. Brain Res. Bull. 49:441–451;1999.

53. Plata-Salaman, C. R.; Sonti, G.; Borkoski, J. P.; Wilson, C. D.;ffrench-Mullen, J. M. H. Anorexia induced by chronic central admin-istration of cytokines at estimated pathophysiological concentrations.Physiol. Behav. 60:867–875; 1996.

54. Porter, M. H.; Arnold, M.; Langhans, W. TNF-a tolerance blocksLPS-induced hypophagia but LPS tolerance fails to prevent TNF-a-induced hypophagia. Am. J. Physiol. 274:R741–R745; 1998.

55. Porter, M. H.; Hrupka, B. J.; Langhans, W.; Schwartz, G. J. Vagal andsplanchnic afferents are not necessary for the anorexia produced byperipheral IL-1beta, LPS, and MDP. Am. J. Physiol. 275:R384–R389;1998.

56. Quan, N.; Whiteside, M.; Herkenham, M. Time course and localizationpatterns of interleukin-1beta messenger RNA expression in brain andpituitary after peripheral administration of lipopolysaccharide. Neuro-science 83:281–293; 1998.

57. Satta, M. A.; Jacobs, R. A.; Kaltsas, G. A.; Grossman, A. B. Endotoxininduces interleukin-1beta and nitric oxide synthase mRNA in rathypothalamus and pituitary. Neuroendocrinology 67:109–116; 1998.

58. Schwartz, G. J.; Plata-Salaman, C. R.; Langhans, W. Subdiaphrag-matic vagal deafferentation fails to block feeding-suppressive effectsof LPS and IL-1 beta in rats. Am. J. Physiol. 273:R1193–R1198; 1997.

452 TURRIN ET AL.

59. Sonti, G.; Ilyin, S. E.; Plata-Salaman, C. R. Anorexia induced bycytokine interactions at pathophysiological concentrations. Am. J.Physiol. 270:R1394–R1402; 1996.

60. Sonti, G.; Ilyin, S. E.; Plata-Salaman, C. R. Neuropeptide Y blocks andreverses interleukin-1b-induced anorexia in rats. Peptides 17:517–520;1996.

61. Turrin, N. P.; Flynn, M. C.; Plata-Salaman, C. R. Neuropeptide Ycounteracts interferon-alpha-induced anorexia. Neuroimmunomodula-tion 6:361–366; 1999.

62. Turrin, N. P.; Plata-Salaman, C. R. Cytokine-cytokine interactions inthe brain. Brain Res. Bull. 51:3–9; 2000.

63. van Dam, A. M.; Poole, S.; Schultzberg, M.; Zavala, F.; Tilders, F. J.

Effects of peripheral administration of LPS on the expression ofimmunoreactive interleukin-1 alpha, beta, and receptor antagonist inrat brain. Ann. N.Y. Acad. Sci. 840:128–138; 1998.

64. Wesche, H.; Neumann, D.; Resch, K.; Martin, M. U. Co-expression ofmRNA for type I and type II interleukin-1 receptors and the IL-1receptor accessory protein correlates to IL-1 responsiveness. FEBSLett. 391:104–108; 1996.

65. Wong, M. L.; Bongiorno, P. B.; Rettori, V.; McCann, S. M.; Licinio,J. Interleukin (IL) 1beta, IL-1 receptor antagonist, IL-10, and IL-13gene expression in the central nervous system and anterior pituitaryduring systemic inflammation: pathophysiological implications. Proc.Natl. Acad. Sci. USA 94:227–232; 1997.

BRAIN CYTOKINES AND LPS 453