Effect of Ciliary Neurotrophic Factor on Body Temperature and Cerebrospinal Fluid Prostanoids in the Cat

Brain Research Bulletin, Vol. 45, No. 1, pp 9 –14, 1998 Copyright © 1998 Elsevier Science Inc. Printed in the USA. All rights reserved 0361-9230/98 $19.00 1 .00

PII S0361-9230(97)00273-6

Effect of Ciliary Neurotrophic Factor on Body Temperature and Cerebrospinal Fluid Prostanoids in the Cat E. S. AKARSU, I. BISHAI AND F. COCEANI Division of Neurosciences, Research Institute, The Hospital for Sick Children, Toronto, Ontario M5G 1X8, Canada [Received 30 April 1997; Accepted 1 July 1997] ABSTRACT: It has been proposed that ciliary neurotrophic factor (CNTF) belongs to the group of cytokines causing fever in response to infectious and inflammatory noxae. The present investigation was undertaken in the conscious cat to verify whether CNTF (human type, hCNTF) is pyrogenic when given either intravenously (IV) or intracerebroventricularly (ICV) and correlate at the same time body temperature with cerebrospinal fluid (CSF) levels of prostaglandin (PG) E2 (i.e., the putative fever mediator in brain) and thromboxane (TX) B2 (the stable TXA2 byproduct) in untreated vs. treated animals. hCNTF (10 mg/kg IV; 1 mg ICV) caused fever by both routes and the increase in body temperature was associated with an upward change in CSF PGE2. Conversely, CSF TXB2 showed no elevation. Similarly unaffected was CSF TXB2 by human interleukin 6 (hIL-6, 1 mg ICV), a cytokine with known pyrogenic and PGE2-promoting actions sharing the signal-transducing mechanism with hCNTF. We conclude that CNTF lends itself to a role in the pathogenesis of fever. The modest PGE2 elevation relatively to other cytokines, specifically hIL-1, is ascribed to the fact that CNTF activates the inducible isoform of arachidonate cyclooxygenase, which is constitutively expressed in brain, without concomitantly promoting the formation of new enzyme. © 1998 Elsevier Science Inc.

sidered of potential benefit to patients with neurologic disorders of a degenerative nature [1,3,14,25]. Among the possible functions of CNTF is the participation in the pathogenetic sequence of fever as an endogenously formed pyrogen [13,15,29]. Consistent with this concept is the pyrogenicity of the agent [29], its ability to elicit hyperthermia and associated acute-phase changes without the involvement of other cytokines [13,15,29], the reversibility of CNTF hyperthermia with indomethacin [29], and the presence in the CNTF receptor of a signal-transducing mechanism (i.e., the gp130 subunit) common to IL-6 [19,28]. IL-6 is well accepted as an endogenous pyrogen [10,12,27] and, according to a recent report [5], would appear to be critical for the onset of fever. Taken together, these facts suggest that CNTF shares with many of the endogenous pyrogens the dependence on prostaglandin (PG) E2 for its central action (see [7]) and, by extension, imply that the response to the compound has the hallmarks of a true fever. Pyrogens may also exert their effect without the involvement of PGE2 [11], but this arrangement is regarded as exceptional. The purpose of our study was to verify whether PGE2 qualifies as a mediator for CNTF fever in brain. An additional aim was to obtain reference data for clinical trials with systemically or intrathecally administered CNTF. Fever is a common side effect of this treatment, particularly when using the systemic route [2]. Accordingly, CNTF was tested in the chronically instrumented cat, both intravenously and intracerebroventricularly, with the intent of reproducing the events during the host response to a febrigenic noxa and the therapeutic use of the compound. Changes in body temperature due to the treatment were correlated with PGE2 levels in the cerebrospinal fluid (CSF). Thromboxane (TX) B2 (the stable TXA2 byproduct) was also measured in experiments with intracerebroventricular hCNTF for a better comparison with results obtained earlier with endotoxin and IL-1 (i.e., the principal exogenous and endogenous pyrogens). Endotoxin and IL-1 have a differential effect on PGE2 vs. TXA2 formation in brain with the former prostanoid being activated throughout the entire febrile episode regardless of the route of administration and the latter being activated only in response to intracerebroventricular administration and transiently [8,9]. Selected experiments were also

KEY WORDS: Fever mechanism, Ciliary neurotrophic factor, Interleukin 6, Prostaglandin E2, Thromboxane.

INTRODUCTION Ciliary neurotrophic factor (CNTF) is a cytosolic peptide forming, together with interleukins 6 and 11 (IL-6 and IL-11) and certain hemopoiesis-regulating factors, a distinct group of cytokines [19,28]. In the central nervous system, CNTF is confined to the astrocytes [4,32]; however, its receptor sites are distributed widely and are particularly abundant on neurons [19,35]. This peptide lacks a signal sequence for release outside cells [33], but its secretion could still occur through a nonconventional mechanism [6,33]. Originally implicated in the growth of neural cells and their survival following injury [24,28], CNTF is now assigned a broader role in the normal operation of neurons, including events that take place at the synapse [6,21–24,34]. In addition, its action is con1

To whom requests for reprints should be addressed.

9

10

AKARSU, BISHAI AND COCEANI

performed with IL-6 to extend the comparison to this CNTFrelated cytokine. MATERIALS AND METHODS Materials Recombinant human CNTF (hCNTF) and recombinant human IL-6 (hIL-6) (both from Regeneron Pharmaceuticals, Tarrytown, NY) were expressed in E. coli. Their biological activity was, respectively, 1.8 ng/ml (ED50 in the embryonic chicken ciliary ganglion neuronal survival bioassay) and 2 pg/ml (ED50 in B9 hybridoma assay). Recombinant proteins lost their pyrogenicity after heating at 70°C for 30 min. Doses were selected from previous studies in our (hIL-6) [12] and other (hCNTF) [29] laboratories and, in the case of hCNTF, were also verified in preliminary tests. hCNTF was dissolved in sterile water (USP; Astra Pharma, Mississauga, ON), as recommended by the supplier, and stored at 270°C. On the day of the experiment, aliquots of this stock solution were diluted with saline (intravenous injection) or Elliott solution (intracerebroventricular injection). hIL-6 was dissolved directly in Elliott solution. Indomethacin was purchased from Sigma (St. Louis, MO) and was dissolved in ethanol (10 mg/ml) for storage. Unbuffered, artificial CSF (Elliott solution) was obtained from Astra. Glassware, syringes, solutions, ventricular implants and their accessories, and intravenous lines were sterile and pyrogen free. Experimental Procedure Experiments were performed on adult cats of either sex (3.5– 4.5 kg body weight). With a single exception, three to five animals were used with each experimental protocol, and protocols were often repeated three or four times in the same animal, taking care to space individual tests appropriately in time (see below). Two animals were used in the remaining protocol (intracerebroventricular vehicle with CSF TXB2 measurement). Following a published procedure [8], animals were cannulated inside the rostral part of the third ventricle and, when required, inside the external jugular vein as well. In one case, the brain cannula was found in the right lateral ventricle (close to the third ventricle) on postmortem examination (see below); however, results did not differ from those of the group and were included in the computation. About 3 weeks were allowed for recovery, and experiments were started after wounds had healed and CSF flowed freely from the ventricle. During tests, cats were not anesthetized and were fitted with thermistor probes coupled to a telethermometer— chart recorder assembly for continuous monitoring of colonic and ear (dorsal surface) temperatures. Skin temperature provided an index of peripheral vasomotor tone. The behaviour of the cats was also noted, and respiratory frequency was counted at regular intervals. Before any intervention, animals were allowed to settle down and attain a steady body temperature (range, 38.1–38.9°C). hCNTF was given both intravenously and intracerebroventricularly (10 mg/kg and 1 mg, respectively), while hIL-6 was given only intracerebroventricularly (1 mg). The volume of intracerebroventricular injectate was 100 ml, while intravenous administration consisted of a 2-ml bolus. CSF was collected at a rate approximating the rate of production (0.12– 0.20 ml in 8 –12 min), and the interval between samples varied from 1 to 4 h, depending on the protocol. In those instances in which animals showed a rise in body temperature, collections were made once or, more frequently, twice at 30-min interval under basal conditions, once or twice through the uprise phase, and after fever had reached a peak. An additional sample was taken late in the course of the fever to intravenous hCNTF

(about 1 h after the initial peak). When animals did not develop fever, CSF was collected at intervals corresponding chronologically to various stages of the fever response. In any experiment, however, the number of CSF collections did not exceed six. Samples were collected into an ice-cold vial containing 2 mg indomethacin and were subsequently stored at 220°C until analysis. At least 5 days elapsed between experiments in the same animal. At the conclusion of the study, the location of the ventricular implant was confirmed by examining the distribution of a dye in the ventricular system [8]. All animal procedures were approved by the Animal Care Committee of The Hospital for Sick Children. Radioimmunoassay PGE2 and TXB2 were assayed directly in CSF samples using radioimmunoassay kits with, respectively, 125I-labelled and 3Hlabelled ligand (DuPont, Boston, MA). Both assay procedures have been validated previously [9,30]. The two prostanoids were measured in samples from different experiments, and the limit of detection was 10 –25 and 100 pg/ml CSF for, respectively, PGE2 and TXB2. Analysis of Data Increases in body temperature are given in the text as the maximal change (DT) from baseline. Data are expressed as means 6 Sem and n refers to the number of experiments. Statistical analysis of unpaired observations was made using the Student’s t-test. Multiple comparisons were made with a one-way analysis of variance for repeated measures either alone (comparison within a group) or with an added test for interaction (comparison between groups). The latter analysis was carried out on log-transformed data to improve normality and a value equal to the limit of detection of the assay was used in those instances in which prostanoids could not be measured in CSF samples. In certain cases, a CSF measurement was missing due to some technical failure and, accordingly, n values are given for each sampling interval. Differences are considered significant for p , 0.05. RESULTS In agreement with previous work [9], PGE2 and TXB2 could be measured in the majority of CSF samples collected at rest (Table 1). With both compounds, concentrations appeared higher in the first of the two controls, though the difference proved to be significant only in the case of PGE2 (Table 1). For this reason, the second set of samples was used as a reference for subsequent treatments. Neither the range of prostanoid concentrations nor the incidence of samples below detection changed with successive experiments and the repeated exposure to the agents under study. Only once did the basal level exceed the above range (164 pg/ml) for no apparent reason, and this value was excluded from the computation. Effects of Intravenous hCNTF Intravenous hCNTF produced a fever in all experiments. As shown in Fig. 1A, body temperature began to rise shortly after the injection (latency, 4 –19 min), attained rapidly a peak (45–99 min), and then started to abate. Complete, or nearly complete, defervescence occurred in 180 –270 min (data not shown). Concomitant with the fever there was an upward trend in the PGE2 content of the CSF, but this elevation did not attain significance at any stage of the response (Fig. 1B). In contrast, no change at all in either body temperature or CSF PGE2 was seen when animals were given the vehicle alone (Fig. 1B).

CNTF FEVER AND CSF PROSTANOIDS

11 TABLE 1

VENTRICULAR CONCENTRATION OF PGE2 AND TXB2 UNDER BASAL CONDITIONS CSF Content First Control

Compound

PGE2 TXB2

Above Detection (pg/ml)

71 6 15 (24) 673 6 315 (9)

Second Control

Below Detection (No. of Samples)

3 6

Above Detection (pg/ml)

37 6 4 (29) 313 6 77 (9)

Below Detection (No. of Samples)

4 6

CSF was collected at a 30-min interval and values are means 6 SEM for No. of samples given in parentheses. In some experiments (see Materials and Methods), there was only one CSF withdrawal and its timing coincided with that of the second sample of a pair. The two control groups are significantly different (p , 0.05) in the case of PGE2. Note that, in comparing the two groups, samples without a measurable compound were assigned a value equal to the limit of detection of the assay (for details, see Materials and Methods).

Effects of Intracerebroventricular hCNTF hCNTF was a pyrogenic agent by the intracerebroventricular route in all, but one, of the five animals tested. In the remaining animal, the change in body temperature was marginal (,0.4°C) and did not exceed the normal circadian fluctuation. When present, the fever differed in pattern from that following intravenous hCNTF. As evident from Fig. 2A, it had a delayed onset (mean, 127; range, 17–157 min) and progressed gradually to a peak (mean, 341; range, 233– 421 min). In addition, the rise in body temperature was sustained, and no reversal was noted during the whole period of recording (max. 8 h). However, as in the case of the fever to the intravenous compound, PGE2 in the CSF increased little and barely reached significance (Fig. 2B). Nonsignificant was instead the change in CSF TXB2 (Fig. 2B1). Vehicle alone, on the other hand, had no effect on both body temperature and prostanoid content of the CSF (Fig. 2). Because it has been previously shown [12] that intracerebroventricular hIL-6, at the same dose as hCNTF, increases body temperature and PGE2 levels in CSF, experiments were now performed to examine hIL-6 action on TXB2. However, despite the occurrence of a fever (DT, 1.08 6 0.17°C; n 5 5), intracerebroventricular hIL-6 had no effect on CSF TXB2, and levels of this prostanoid remained below detection or fluctuated within a normal range of measurable values (100 –320 pg/ml). By either route, hCNTF did not lose its effectiveness as a pyrogen on repeated administrations to the same animal, nor did the animal itself show obvious sequelae from exposure to the agent. Specifically, loss of body weight has been reported as a complication to any intensive treatment in animals [18,21] and humans [1,25]; however, it was not observed under the conditions of our study. Otherwise, in their immediate response to hCNTF, animals showed a modest increase in the respiratory rate (from 20 –24 to 30 –39 breaths/min) and the typical signs of a developing fever (i.e., shivering, reduced skin temperature, sedation). DISCUSSION This study confirms that hCNTF is a pyrogenic agent, which, in common with other cytokines, specifically hIL-1 and hIL-6, has the property of eliciting fever with a different time course by the intravenous vs. the intracerebroventricular route [9,12]. Its potency is low relative to hIL-1 [9], but approximates that of hIL-6 [12]. Contrary to hIL-1 [9], hCNTF causes only a modest rise in CSF PGE2 by either route, and this effect does not reach significance or is barely significant. hIL-6, on the other hand, is a more effective

FIG. 1. Conscious cat. Record of body temperature (A) and content of PGE2 in third ventricular CSF (B) under basal conditions and after intravenous bolus injection (at arrow) of hCNTF (10 mg/kg) (closed symbols and bars) or saline (open symbols and bars). Temperature shown is change from baseline (time 5 0 min). Number of experiments appears with the temperature recording and above each set of CSF values. Note that the CSF control group for the hCNTF series does not include one sample in which PGE2 content exceeded greatly the norm (164 pg/ml).

12

FIG. 2. Conscious cat. Record of body temperature (A) and content of PGE2 (B) and TXB2 (B1) in third ventricular CSF under basal conditions and after intracerebroventricular bolus injection (at arrow) of hCNTF (1 mg) (closed symbols and bars) or Elliott solution (open symbols and bars). In certain experiments where PGE2 was assayed, CSF was withdrawn twice during fever uprise and the mean of the two values was used for computation. Note that in one animal the sampling cannula was located in the lateral instead of the third ventricle (see text); however, findings did not differ from those of the group and were pooled. Number of experiments appears with the temperature recording and above each set of CSF values. * p , 0.05 from control.

PGE2 stimulant than hCNTF, though it is still weak compared to hIL-1 [12]. PGE2 values, whether at rest or in the course of the hCNTF fever, do not change through successive experiments in the same animal, which argues against the presence of any reactive alteration in the glia consequent to the treatment [20,36]. Peculiarly, both hCNTF and hIL-6 differ from hIL-1 [8] in not being able to stimulate TXA2 synthesis when given intracerebroventricularly. From this premise, our discussion will concern the following

AKARSU, BISHAI AND COCEANI three issues: the reason for the weak PGE2 activation by hCNTF in the face of a full-fledged fever with the related question of why TXA2 is not affected at all when this cytokine, or the allied hIL-6, is delivered directly onto the brain; the feasibility of CNTF functioning as an endogenous pyrogen; and the implications of our findings for current and prospective clinical trials employing hCNTF. The apparent incongruity between changes in body temperature and brain PGE2 synthesis in the hCNTF-treated cat unlikely reflects a methodological problem inherent to accessibility, location, or susceptibility of target sites(s). Results were virtually identical regardless of the route, blood vs. CSF, taken by the compound to reach the tissue of the brain and, moreover, they did not differ between the first and subsequent tests in the same animal. In addition, CNTF receptors are known to be distributed diffusely throughout the brain [19,35]. Incompetence of the human material for the cat can also be ruled out because hCNTF is effective across species and, more importantly, has been proven competent under the conditions of our study, at least insofar as its effect on body temperature is concerned. It is, therefore, reasonable to conclude that the modest PGE2 stimulation is a true finding, which in itself does not necessarily exclude an intermediary role of the compound in the pathogenetic sequence of the fever. Consistent with this conclusion is the efficacy of indomethacin as an antipyretic agent against hCNTF [29]. In attempting to explain the peculiarly weak activation of PGE2 by hCNTF, a possible clue may be found in recent studies of the prostaglandin synthetic system (i.e., the cyclooxygenase system, COX) and our own studies of hIL-6. Of relevance here is the fact that brain, unlike most other organs, may express the inducible isoform, COX2, constitutively [31], and that hIL-6 differs from hIL-1 in not being able to induce COX2 [26]. Conceivably, hCNTF behaves as hIL-6 in the latter respect. An additional difference between hIL-6 and hIL-1 emerging from our previous work is that the former may stimulate PGE2 synthesis in a tissue-specific manner. We have shown, for example, that hIL-6 is effective on brain but not on blood mononuclear cells [12]. Significantly, the same cells are also unresponsive to hCNTF [29]. Taken together, these data enable us to formulate a possible scheme on the cytokine-COX interaction. Central to our proposal is the notion that cytokines exert their effect via COX2 [31] and that, by lacking the ability to induce this enzyme, hIL-6 and hCNTF can only affect tissues where COX2 is expressed constitutively. From this one may infer that mononuclear cells fail to respond to hIL-6 and hCNTF due to the absence of COX2 [17] and that in brain the same cytokines are less potent PGE2 stimulators than hIL-1 because they act upon COX2 without concomitantly promoting the expression of new enzyme. This arrangement may also explain the differential action of cytokines on brain TXA2 synthesis, with hIL-1 being an activator [8] and hIL-6 and hCNTF being conversely without effect (this study). Extending the above speculation further one could surmise that acceleration of TXA2 synthesis takes place only through upregulation of COX2 and the ensuing recruitment of a separate arachidonate pool. An apparent weakness in our reasoning is that hIL-6 and hCNTF differ in their effectiveness as PGE2 stimulators despite their expectedly similar action on the COX system. It must be stressed, however, that the difference is not major (compare with [12]) and may reflect some diversity in the receptor/ transduction step rather than in the COX effector. In the end, with this scheme we assume that, notwithstanding the modest change in CSF, the PGE2 stimulation brought about by hCNTF occurs at some place in brain, which is critical for the genesis of fever. Coincidentally, we are able to resolve the apparent

CNTF FEVER AND CSF PROSTANOIDS paradox of systemically administered endotoxin and hIL-1 activating PGE2 selectively [9] despite the induction of IL-6 within the confines of brain [10]. The question of whether CNTF functions as an endogenous pyrogen in concert with other cytokines remains open. Ordinarily, CNTF is not present in either blood [2] or CSF [1]. It becomes measurable in the blood of patients with septic shock [16]; however, under the conditions of the assay it could not be ascertained whether this elevation is sustained by a secretory process or the release from damaged cells. Indeed, there are uncertainties on the actual capability of cells to secrete CNTF [6,33]. In addition, the CNTF receptor may be activated by other ligand(s) [19], and this possibility introduces a further complication in defining a role for the compound. Translated to the clinical situation, our work confirms that fever should be expected as a side effect of the hCNTF treatment. This also applies to any future trial in which the compound is delivered directly onto the brain. The intrathecal route is now considered a better alternative to the systemic route for efficacy and general safety [1,14]. In summary, our study proves that hCNTF is a pyrogenic agent and implicates PGE2 as its messenger on central thermoregulatory pathways. With the approach used it is not possible to determine whether these findings are applicable to the endogenously formed compound and whether, ultimately, CNTF can be regarded as a full-fledged contributor to the host defense reaction. Fever, on the other hand, must be viewed as a potential complication of any therapeutic intervention with hCNTF.

13

9.

10.

11.

12.

13.

14.

15.

16.

17. ACKNOWLEDGEMENTS

This work was supported by the Medical Research Council of Canada and a grant from Regeneron Pharmaceuticals. The assistance of R. B. Cavalcanti in part of the experiments is gratefully acknowledged. E.S.A. was supported in part by a NATO Science Fellowship that had been awarded through the Scientific Research Council of Turkey (TUBITAK).

18. 19.

20.

REFERENCES 1. Aebischer, P.; Schluep, M.; De´glon, N.; Joseph, J.-M.; Hirt, L.; Heyd, B.; Goddard, M.; Hammang, J. P.; Zurn, A. D.; Kato, A. C.; Regli, F.; Baetge, E. E. Intrathecal delivery of CNTF using encapsulated genetically modified xenogenic cells in amyotrophic lateral sclerosis patients. Nat. Med. 2:696 – 699; 1996. 2. ALS CNTF Treatment Study (ACTS) phase I–II Study Group. The pharmacokinetics of subcutaneously administered recombinant human ciliary neurotrophic factor (rHCNTF) in patients with amyotrophic lateral sclerosis: Relation to parameters of the acute-phase response. Clin. Neuropharmacol. 18:500 –514; 1995. 3. ALS CNTF Treatment Study Group. A double-blind placebo-controlled clinical trial of subcutaneous recombinant human ciliary neurotrophic factor (rHCNTF) in amyotrophic lateral sclerosis. Neurology 46:1244 –1249; 1996. 4. Carroll, P.; Sendtner, M.; Meyer, M.; Thoenen, H. Rat ciliary neurotrophic factor (CNTF): Gene structure and regulation of mRNA levels in glial cell cultures. Glia 9:176 –187; 1993. 5. Chai, Z.; Gatti, S.; Toniatti, C.; Poli, V.; Bartfai, T. Interleukin (IL)-6 gene expression in the central nervous system is necessary for fever response to lipopolysaccharide or IL-1b: A study on IL-6-deficient mice. J. Exp. Med. 183:311–316; 1996. 6. Clatterbuck, R. E.; Price, D. L.; Koliatsos, V. E. Ciliary neurotrophic factor stimulates the expression of glial fibrillary acidic protein by brain astrocytes in vivo. J. Comp. Neurol. 369:543–551; 1996. 7. Coceani, F. Prostaglandins and fever: Facts and controversies. In: Mackowiak, P., ed. Fever: Basic mechanisms and management. New York: Raven Press; 1991:59 –70. 8. Coceani, F.; Bishai, I.; Dinarello, C. A.; Fitzpatrick, F. A. Prostaglan-

21.

22.

23.

24. 25.

26.

27.

28. 29.

din E2 and thromboxane B2 in cerebrospinal fluid of afebrile and febrile cat. Am. J. Physiol. 244:R785–R793; 1983. Coceani, F.; Lees, J.; Bishai, I. Further evidence implicating prostaglandin E2 in the genesis of pyrogen fever. Am. J. Physiol. 254:R463– R469; 1988. Coceani, F.; Lees, J.; Mancilla, J.; Belizario, J.; Dinarello, C. A. Interleukin-6 and tumor necrosis factor in cerebrospinal fluid: Changes during pyrogen fever. Brain Res. 612:165–171; 1993. Davatelis, G.; Wolpe, S. D.; Sherry, B.; Dayer, J.-M.; Chichepostiche, R.; Cerami, A. Macrophage inflammatory protein-1: A prostaglandinindependent endogenous pyrogen. Science 243:1066 –1068; 1989 Dinarello, C. A.; Cannon, J. G.; Mancilla, J.; Bishai, I.; Lees, J.; Coceani, F. Interleukin-6 as an endogenous pyrogen: Induction of prostaglandin E2 in brain but not in peripheral blood mononuclear cells. Brain Res. 562:199 –206; 1991. Dittrich, F.; Thoenen, H.; Sendtner, M. Ciliary neurotrophic factor: Pharmacokinetics and acute phase response in rat. Ann. Neurol. 35: 151–163; 1994. Emerich, D. F.; Lindner, M. D.; Winn, S. R.; Chen, E.-Y.; Frydel, B. R.; Kordower, J. H. Implants of encapsulated human CNTF-producing fibroblasts prevent behavioural deficits and striatal degeneration in a rodent model of Huntington’s disease. J. Neurosci. 16:5168 – 5181; 1996. Fantuzzi, G.; Benigni, F.; Sironi, M.; Conni, M.; Carelli, M.; Cantoni, L.; Shapiro, L.; Dinarello, C. A.; Sipe, J. D.; Ghezzi, P. Ciliary neurotrophic factor (CNTF) induces serum amyloid A, hypoglycaemia and anorexia, and potentiates IL-1 induced corticosterone and IL-6 production in mice. Cytokine 7:150 –156; 1995. Guillet, C.; Fourcin, M.; Chevalier, S.; Pouplard, A.; Gascan, H. ELISA detection of circulating levels of LIF, OSM and CNTF in septic shock. Ann. NY Acad. Sci. 762:407– 409; 1995. Hempel, S.L.; Monick, M.M.; Hunninghake, G. W. Lipopolysaccharide induces prostaglandin H synthase-2 protein and mRNA in human alveolar macrophages and blood monocytes. J. Clin. Invest. 93:391– 396; 1994. Henderson, J. T.; Mullen, B. J. M.; Roder, J. C. Physiological effects of CNTF-induced wasting. Cytokine 8:784 –793; 1996. Ip, N. Y.; Yancopoulos, G. D. The neurotrophins and CNTF: Two families of collaborative neurotrophic factors. Am. Rev. Neurosci. 19:491–515; 1996. Kahn, M. A.; Ellison, J. A.; Speight, G. J.; de Vellis, J. CNTF regulation of astrogliosis and the activation of microglia in the developing rat central nervous system. Brain Res. 685:55– 67; 1995. Kaira, S. P.; Xu, B.; Dube, M.; Kaibara, D.; Martin, D.; Moldawer, L. L. Ciliary neurotrophic factor is a potent anorectic cytokine acting through hypothalamic neuropeptide. Y. Proc. Soc. Neurosci. Meeting 22:1352; 1996. Kew, J. N. C.; Sofroniew, M. V. Ciliary neurotrophic factor supports p75NGFR-immunoreactive non-cholinergic, but not cholinergic, developing septal neurons in vitro. Neuroscience 66:793– 804; 1995. Lesser, S. S.; Lo, D. C. Regulation of voltage-gated ion channels by NGF and ciliary neurotrophic factor in SK-N-SH neuroblastoma cells. J. Neurosci. 15:253–261; 1995. Lo, D. C. A central role for ciliary neurotrophic factor? Proc. Natl. Acad. Sci. USA 90:2557–2558; 1993. Miller, R. G.; Petajan, J. H.; Bryan, W. W.; Armon, C.; Barohn, R. J.; Goodpasture, J. C.; Hoagland, R. J.; Parry, G. J.; Ross, M. A.; Stromatt, S. C.; and the rhCNTF ALS Study Group. A placebocontrolled trial of recombinant human ciliary neurotrophic (rhCNTF) factor in amyotrophic lateral sclerosis. Ann. Neurol. 39:256 –260; 1996. Murakami, M.; Matsumoto, R.; Austen, K.F.; Arm, J.P. Prostaglandin endoperoxide synthase-1 and -2 couple to different transmembrane stimuli to generate prostaglandin D2 in mouse bone marrow-derived mast cells. J. Biol. Chem. 269:22269 –22275; 1994. Rothwell, N. J.; Luheshi, G.; Toulmond, S. Cytokines and their receptors in the central nervous system: Physiology, pharmacology and pathology. Pharmacol. Ther. 69:85–95; 1996. Sendtner, M.; Carroll, P.; Holtmann, B.; Hughes, R. A.; Thoenen, H. Ciliary neurotrophic factor. J. Neurobiol. 25:1436 –1453; 1994. Shapiro, L.; Zhang, X.-X.; Rupp, R. G.; Wolff, S. M.; Dinarello, C. A.

14

30. 31. 32.

33.

AKARSU, BISHAI AND COCEANI Ciliary neurotrophic factor is an endogenous pyrogen. Proc. Natl. Acad. Sci. USA 90:8614 – 8618; 1993. Sirko, S.; Bishai, I.; Coceani, F. Prostaglandin formation in the hypothalamus in vivo: Effect of pyrogens. Am. J. Physiol. 256:R616 –R624; 1989. Smith, W. L.; Garavito, R. M.; DeWitt, D. L. Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J. Biol. Chem. 271: 33157–33160; 1996. Sto¨ckli, K. A.; Lillien, L. E.; Na¨her–Noe´, M.; Breitfeld, G.; Hughes, R. A.; Raff, M. C.; Thoenen, H. Regional distribution, developmental changes, and cellular localization of CNTF-mRNA and protein in the rat brain. J. Cell Biol. 115:447– 459; 1991. Sto¨ckli, K. A.; Lottspeich, F.; Sendtner, M.; Masiakouski, P.; Carroll, P.; Go¨tz, R.; Lindholm, D.; Thoenen, H. Molecular cloning, expression

and regional distribution of rat ciliary neurotrophic factor. Nature 343:920 –923; 1989. 34. Stoop, R.; Poo, M.-m. Potentiation of transmitter release by ciliary neurotrophic factor requires somatic signaling. Science 267:695– 699; 1995. 35. Watanabe, D.; Yoshimura, R.; Khalil, M.; Yoshida, K.; Kishimoto, T.; Taga, T.; Kiyama, H. Characteristic localization of gp130 (the signaltransducing receptor component used in common for IL-6/IL-11/ CNTF/LIF/OSM) in the rat brain. Eur. J. Neurosci. 8:1630 –1640; 1996. 36. Winter, C. G.; Saotome, Y.; Levison, S. W.; Hirsh, D. A role for ciliary neurotrophic factor as an inducer of reactive gliosis, the glial response to central nervous system injury. Proc. Natl. Acad. Sci. USA 92:5865– 5869; 1995.