Interleukin-1 in the Brain

6:51–93.

1

Location of interleukin-1 in the nervous system

MARIANNE SCHULTZBERG,     Clinical Research Center, Division of Medical Cell and Neurobiology, Karolinska Institute, Huddinge Hospital, S-141 86 Huddinge, Sweden

Publisher Summary

An important activity related to interleukin-1 (IL-1) is its proliferative effects, especially in the nervous system. This chapter discusses the morphological data on the occurrence of IL-1 in the peripheral and central nervous system with special reference to the role of IL-1 in the hypothalamo-pituitary-adrenocortical axis and as a neuronal growth factor. One of the original actions of EL-1 that lead to its characterization is the pyrogenic effect. Electrophysiological studies have demonstrated that IL-1 stimulates neurons in the anterior hypothalamus where the temperature set point is regulated. Hence, the peripheral administration of a well-defined protein may directly or indirectly affect temperature set point centres in the brain. In an in vitro system based on mouse spinal cord and dorsal root ganglion cells, it was observed that addition of antibodies to IL-1α resulted in an increase in nerve cell degeneration, and this effect was blocked by IL-1α. Also, the addition of vasoactive intestinal polypeptide (VIP) blocked the neurodegenerative effects of IL-1 antibodies.

The interrelationship between the nervous and immune systems has, in recent years, attracted considerable attention and been the subject of an increasing number of studies. In order to begin to understand the morphological and molecular basis for this interaction, we initiated investigations on the possible occurrence of an immune system messenger, interleukin-1 (IL-1), in the nervous system. This cytokine had been shown to have several effects on the nervous system, such as inducing fever (Stitt, 1981; Bligh, 1982; Dascombe, 1985) and slow wave sleep (Krueger et al., 1984; Tobler et al., 1984). Furthermore, IL-1 was chosen because it is one of the first substances to be released upon infection and inflammation (see Dinarello, 1984). There are also several studies that strongly indicate a role for IL-1 in the hormonal interactions between the hypothalamus, the pituitary and the adrenal gland. Both the glucocorticoid and gonadal hormone secretion is affected by administration of IL-1 (see Dunn, 1990). Another important activity related to IL-1 is its proliferative effects, especially in the nervous system. It has been shown to stimulate glial proliferation (cf. Nieto-Sampedro and Berman, 1987), particularly of astrocytes, and has also been demonstrated to induce translation and transcription of nerve growth factor (NGF) (Lindholm et al., 1987). Induction by IL-1 of NGF and NGF mRNA has been shown to occur both in a peripheral nerve upon sectioning, and in the central nervous sytem (Spranger et al., 1990).

In the following, a description of the morphological data on the occurrence of IL-1 in the peripheral and central nervous system is presented, with special reference to the role of IL-1 in the hypothalamo-pituitary-adrenocortical axis and as a neuronal growth factor.

1.1 IL-1 in the peripheral nervous system

Immunohistochemical studies using antisera raised against synthetic peptides (IL-1α169-194 and IL-1α201-215) of the murine IL-1α precursor protein (Lomedico et al., 1984) revealed the occurrence of IL-1 immunoreactive material mainly in the peripheral nervous system (Schultzberg et al., 1987). A general finding was the occurrence of varicose, IL-1 immunoreactive fibres around blood vessels, in close relation to the vascular wall (Figs. 1.1C, 1.2B). The appearance and distribution of these fibres strongly resembles the noradrenergic innervation of blood vessels. Double staining experiments indicated that the IL-1 immunoreactive material coexists with neuropeptide Y (NPY) (Schultzberg et al., in preparation), a neuropeptide which is known to occur in noradrenergic sympathetic neurons (Lundberg et al., 1982). The possibility that IL-1 released from nerve fibres innervating blood vessels has a vasoactive action has yet to be demonstrated. However, IL-1 has been shown to stimulate proliferation of vascular smooth muscle cells (Bonin et al., 1989).

Fig. 1.1A–E Immunofluorescence micrographs of sections of the rat adrenal gland (A, B), urinary bladder (C) and rat PC12 cells (D, E), after incubation with antiserum to mIL-1α169-194 (A, C), PNMT (B) and holo mIL-1α (D). IL-1 immunoreactive cells are observed in the adrenal medulla (A), corresponding to PNMT-negative (B) and hence noradrenergic chromaffin cells (arrows in A and B). Intense immunoreaction to IL-1 is seen in the NGF-treated PC12 cells (D). Cells incubated with control serum lacked this immunoreaction (E). Numerous IL-1 immunoreactive nerve fibres can be seen in the smooth muscle layers (sm), and surrounding both an artery (a) and vein (V) in the wall of the urinary bladder (C).

Fig. 1.2 A–E Immunofluorescence micrographs of sections of the rat colon (A, D), spleen (B), hypothalamus (C) and pituitary gland (E), after incubation with antiserum to mIL-1α169-194 (A-D) and a synthetic peptide of the IL-1 receptor (E). Abundant IL-1 immunoreactive nerve fibres can be seen in the mucosa (A), smooth muscle layers (D) and enteric plexuses (arrow in D) of the colon. Note fibres around the basis of the glands (gl) in the mucosa and in the lamina muscularis mucosa (1mm) (A). Many IL-1 positive fibres are observed around blood vessels both in the red and white pulp (arrow in B) of the spleen (b trabecular vein). A few weakly fluorescent neurons are seen in the anterior hypothalamic nucleus (arrows in C). Short, varicose fibres are emanating from the cell bodies. A dense IL-1 immunoreaction is noted in the intermediate lobe of the pituitary gland (E).

Location of IL-1 immunoreactive nerve fibres around blood vessels was also noted in the lymphatic organs (Schultzberg et al., 1987). These fibres were particularly abundant around blood vessels supplying the spleen (Fig. 1.2B), both in the red and white pulp. It is possible that some of the IL-1 positive fibres in the spleen and other lymphatic organs are connecting to immunocompetent cells, similar to the findings by Bulloch (1985) and Feiten et al. (1985) with regard to the cholinergic, noradrenergic and peptidergic innervation.

IL-1 immunoreactive nerve fibres also occur in other tissues. Many varicose fibres can be seen in the smooth muscle layers of the gastrointestinal tract and the urogenital tract (Figs 1.1C, 1.2D), and there are also fibres surrounding the nerve cell bodies in the myenteric and submucous plexus of the digestive tract (Fig. 1.2D) (Schultzberg et al., 1987). These fibres may represent postganglionic sympathetic fibres where noradrenaline and IL-1 are thus colocalized. Alternatively, they may originate in intrinsic myenteric and submucosal ganglion cells, where IL-1 would coexist with acetylcholine and a number of neuropeptides (see Furness and Costa, 1980; Schultzberg et al., 1980). In the thick muscular wall of the vas deferens, the distribution and appearance of the IL-1 positive fibres show a strong resemblance to the noradrenergic innervation of this tissue. To our knowledge there is no information on the possible action of IL-1 in these sites. It is not known whether IL-1 acts on, for example, the muscle layers in the gut by inducing contraction or relaxation, either directly on the smooth muscle or via actions on intrinsic or extrinsic enteric innervation. IL-1 released from nerves in the gut should also be considered as a possible factor in inflammatory diseases of this organ. Furthermore, IL-1 immunoreactive fibres were observed in the gastrointestinal mucosa, where some fibres surround the glands (Fig. 1.2A). It is interesting to note that IL-1 may affect mucosal secretion (Han et al., 1987).

Nerve fibres immunoreactive to IL-1 have recently been observed in rat long bones (Bjurholm et al., 1991). These fibres are located predominantly in the vicinity of, or within, blood vessel walls, both in epiphyseal and diaphyseal bone, including bone marrow, periost and adjoining connective tissue. However, IL-1 immunoreactive fibres were also frequently observed in the epiphyseal growth plate near the cartilage, both in close relation to blood vessels and without apparent vascular relationship. This distribution closely resembles the autonomic innervation described previously (Bjurholm et al., 1988). The occurrence of intraosseal nerves containing IL-1-like immunoreactivity is interesting in view of its potent effects on bone. Both inhibitory and stimulatory actions of IL-1 have been demonstrated in osteoclasts as well as osteoblasts (Gowen et al., 1983; Ikeda et al., 1988; Stashenko et al., 1987).

1.2 IL-1 and the hypothalamo-pituitary-adrenocortical axis

Administration of IL-1 stimulates an increase in plasma levels of both corticotropin releasing factor (CRF) and adrenal corticotrophic hormone (ACTH) (Berkenbosch et al., 1987; Bernton et al., 1987; Sapolsky et al., 1987; Uehara et al., 1987). It is therefore of interest to note the occurrence of large amounts of IL-1α in the adrenal gland (Schultzberg et al., 1989). This material is localized in the noradrenergic chromaffin cells (Fig. 1.1 A, B; 1.3 A, B) in both rat and mouse, and in a recent study the production of IL-1α in a cell line, rat pheochromocytoma (PC12), was demonstrated (Fig. 1.1D, E) (Alheim et al., 1991). More recent studies have shown that also IL-1β is present in the adrenal chromaffin cells (see Andersson-Fisone, 1991). It has also been shown that lipopolysaccharides (LPS) induce a marked increase in IL-1α in the rat adrenal gland and the PC12 cells (Bartfai et al., 1990; Alheim et al., 1991). The amounts of IL-1 in the rat adrenal gland, as measured by ELISA or RIA, are considerable, and if all was released it would yield circulatory levels of about 10−12 M. It can so far only be speculated that IL-1 in the chromaffin cells may constitute a pool which upon release by nerve activity or stress could help to maintain high circulatory levels of IL-1 to balance the immune response. In support of a release of IL-1 from the adrenal gland are the findings that administration of cholinergic agonists such as nicotine and carbachol cause a marked decrease in IL-1α protein in the gland, whereas IL-1α mRNA was increased (Andersson et al., 1992). The latter effect may reflect the cells’ attempts to increase biosynthesis in order to reconstitute the intracellular levels of IL-1. IL-1 secreted from the adrenal medulla may serve as an immunosuppressant via its actions on the hypothalamo-pituitary-adrenocortical axis. In this case, IL-1-induced stimulation of CRF and ACTH release would result in increased circulating levels of corticosterone which has a well-known immunosuppressive action.

Fig. 1.3 A–B Immunofluorescence micrographs of consecutive sections of the mouse adrenal gland after incubation with antiserum to mIL-1169-194 (A) and PNMT (B). Note the correspondence between IL-1 immunoreactive chromaffin cells (arrows in A) and PNMT-negative cells (arrows in B). Asterisk indicates the same blood vessel.

In order to further investigate this possibility, it is important to localize the receptors for IL-1, and to analyze the regulation of IL-1 receptors in this system. IL-1 receptor autoradiography in the rat has proven to be difficult, at least in our hands, with the tools available. There is one receptor autoradiographic study demonstrating the overall distribution of IL-1α receptor binding in the rat brain (Farrar et al., 1987), including dense binding in, for example, the hypothalamus and hippocampus. More recently, a study of IL-1α and β receptors in homogenates and in sections of the mouse brain indicated that the highest density of binding occurs in the dentate gyrus of the hippocampus (Takao et al., 1990; Haour et al., 1990). Using a different approach, antisera raised against synthetic peptides of the human IL-1 receptor protein were used in immunohistochemical studies (Schultzberg et al., 1991). Immunoreactivity was observed in the rat and mouse pituitary gland with two antisera, directed at different regions of the IL-1 receptor. The immunoreactive material was encountered in a population of cells in the anterior lobe and in the cells of the intermediate lobe (Fig. 1.2E). This distribution pattern agrees with a stimulatory effect of IL-1 on ACTH release from the pituitary, but further characterization is required since there are conflicting data on the exact site of action of IL-1 in the regulation of the glucocorticoid release. Thus, it is conceivable that both the hypothalamus and the pituitary gland are involved in the response to IL-1 on ACTH release.

1.3 IL-1 in the central nervous system

The source of IL-1 acting in different regions of the brain has long been a matter of debate. One of the original actions of IL-1 that lead to its characterization is the pyrogenic effect (see Dinarello, 1984). Dinarello and coworkers showed that intraperitoneal administration of human recombinant IL-1b to endotoxin-resistant mice causes fever. Electrophysiological studies had earlier demonstrated that IL-1 stimulates neurons in the anterior hypothalamus where the temperature set point is regulated (Bligh, 1982; Dascombe, 1985; Stitt, 1981). Hence, the peripheral administration of a well-defined protein may directly or indirectly affect temperature set point centres in the brain. The source of IL-1 acting at these sites in the brain is not yet established. However, several possibilities have been suggested. One is that IL-1 passes the blood brain barrier via circumventricular organs, such as the organum vasculosum lamina terminale (OVLT). In a study by Banks and collaborators (Banks et al., 1989), results were presented indicating that IL-1 may be actively transported both into and out of the brain. Alternatively, there is also a clear possibility that there is an endogenous synthesis of IL-1 in the brain. An early study on human postmortem brain tissue demonstrated IL-1β immunoreactive nerve cell bodies and fibres in the hypothalamus (Breder et al., 1988). More recently, the distribution of IL-1b immunoreactive neurons in the rat brain was described (Lechan et al., 1990). In our laboratory, a few weakly immunofluorescent neuronal cell bodies and nerve fibres could be seen with antibodies directed towards IL-1α. Interestingly, these neurons mainly occurred in the anterior hypothalamic nucleus (Fig.1.2C) (Schultzberg, 1991). IL-1b-positive neurons have been found in the magnocellular part of the paraventricular hypothalamic nucleus (Lechan et al., 1990). Particularly dense networks of fibres were observed around the pyramidal cells in the CA3 of the hippocampus, and in the olfactory tubercle. In contrast to the immunohistochemical results discussed above, in situ hybridization histochemical studies showed the occurrence of IL-1b mRNA predominantly in the pyramidal cell layer of the hippocampus (Bandtlow et al., 1990). Thus, there is either a transcription of IL-1 DNA in the pyramidal cells without translation into protein, or the translation is minimal during normal conditions, and therefore does not permit the demonstration of IL-1 by immunocytochemical techniques. Alternatively, IL-1b mRNA visualized by in situ hybridizations is in fact contained within the nerve terminals stained for IL-lβ by immunohistochemistry. The possibility of mRNA occurring in nerve terminals was recently addressed by Jirikowski et al. (1990).

The localization of IL-1α mRNA in the pyramidal cell layer of the hippocampus is notable also in view of the demonstration of NGF mRNA in this layer (Ayer-LeLievre et al., 1988; Rennert and Heinrich, 1986; Whittemore et al., 1988). Interestingly, it was shown that IL-1 gives rise to an induction of NGF in the hippocampus which was stronger than the NGF induction observed in cultured astrocytes (Spranger et al., 1990). Therefore, it was suggested that the pyramidal cells constitute the main target of IL-1 induced NGF production in the hippocampus. Whether or not IL-1 is released from these neurons or nerve terminals innervating these neurons upon injury is at present unknown. However, increased levels of IL-1 mRNA as estimated by Northern blot analysis were observed in the hippocampus after injection of kainic acid (Minami et al., 1990), an excitotoxic amino acid known to cause selective neuronal degeneration in the hippocampus (see Coyle et al., 1981; Schwarcz et al., 1987). The exact localization of the increased IL-1 mRNA transcription requires further morphological analysis by, for example, in situ hybridization histochemistry. It has been shown earlier that levels of IL-1 are increased by mechanical injury to the brain (Giulian and Lachman, 1985). Results from in vitro studies indicate that IL-1 is produced by microglia (Giulian et al., 1986), and it has been suggested that amoeboid microglia accumulating at the site of injury is the source of increased IL-1 levels.

In summary, considerable data indicate that IL-1 is involved in the stimulation of growth and regeneration in the nervous system. In an in vitro system based on mouse spinal cord and dorsal root ganglion cells it was observed that addition of antibodies to IL-1α resulted in an increase in nerve cell degeneration (Brenneman et al., 1991), and this effect was blocked by IL-1α. Also the addition of vasoactive intestinal polypeptide (VIP) blocked the neurodegenerative effects of IL-1 antibodies. A possible role for IL-1 to either directly or indirectly act as a survival factor was