Differential induction of interleukin-1β and tumour necrosis factor-α may account for specific patterns of leukocyte recruitment in the brain

Abstract

In peripheral tissue, IL-1β has been shown to induce TNFα expression and vice versa, resulting in mixed neutrophil and mononuclear cell recruitment to the site of injury. This has led to the concept of crosstalk in peripheral cytokine signalling pathways. In the brain parenchyma, however, restricted patterns of leukocyte recruitment following the focal injection of pro-inflammatory agents into the brain are observed. This study investigates the expression of the principal pro_inflammatory cytokines—IL-1β and TNFα—in the brain after IL-1β, TNFα, NMDA or endotoxin injection into the brain parenchyma of rats. Each of these agents gives rise to a distinct pattern of acute leukocyte recruitment at 24 h. We found that IL-1β induces de novo synthesis of additional IL-1β but not TNFα, as determined by RT-PCR and ELISA, and TNFα does not induce either itself or IL-1β. Injection of NMDA results in IL-1β, but not TNFα up-regulation. Injection of IL-1β or NMDA is associated with neutrophil recruitment whereas injection of TNFα is associated with mononuclear cell recruitment. Following injection of endotoxin, both TNFα and IL-1β levels are elevated and neutrophils and mononuclear cells are recruited to the brain. These data suggest that the signalling pathways that are present in the periphery are modified in the brain and that differential induction of TNFα and IL-1β may have a role in the atypical pattern of leukocyte recruitment observed in the brain.

Introduction

An early step in an acute inflammatory response to injury or disease in peripheral tissues is the release of cytokines followed by rapid cellular recruitment [37]. The kinetics of cytokine release and the cellular responses involved in acute inflammation have been extensively investigated in peripheral tissues, and similar network relationships have been assumed to exist among the principal pro-inflammatory cytokines in the brain. Indeed, the term ‘cytokine network’ has been widely used to describe cytokine biology in the brain [2], [61]. However, this network has never been studied systematically in vivo.

Cytokines such as tumour necrosis factor-α (TNFα) and interleukin-1β (IL-1β) are of primary importance in the initiation and propagation of the acute phase of the inflammatory response. In peripheral tissue, exogenously administered endotoxin, IL-1β and TNFα each give rise to a similar pattern of cytokine induction, involving an autocrine self-augmentation and the expression of additional cytokines resulting in similar histopathology not specific to the initial challenge [15], [21], [29]. For example, in the skin IL-1β induces TNFα; TNFα induces IL-1β, and there are similar mixed leukocyte recruitment profiles of both monocytes and neutrophils following either injection [37]. Cytokines contribute to inflammatory processes in the brain and participate in the pathology of neurological disease associated with inflammation [23], [45], [64]. Endogenous cytokine production in the brain plays additional roles in mediating the febrile response [26], the activation of the hypothalamic–pituitary–adrenal axis, and other sickness-related behaviours [9], [17], [61]. Increased production of inflammatory cytokines in the brain has been reported in brain trauma [39], [56], as well as in acute bacterial infections of the brain [25]. However, our understanding of the role of the principal inflammatory mediators in the brain remains incomplete.

To determine the effect of cytokines in the brain, several groups have directly injected cytokines, or endotoxin inducing IL-1β and TNFα, into various sites within the brain [38], [47], [48], [49], [50], [54]. These results have clearly demonstrated that the nature of the inflammatory response that ensues is site-dependent. When injection sites include the meninges, the ventricles [31], [41], [49] or the eye [13], [38], the response resembles that found in peripheral tissues (reviewed in Ref. [12]). However, if the injection is made into the brain parenchyma [3], the response is altered with respect to the kinetics and the nature of the cellular infiltrate [1], [3], [7], [34]. This atypical inflammatory response is developmentally acquired, since injection of endotoxin or IL-1β into the brain parenchyma of immature rodents [3], [34] elicits a more florid response than that seen in adults, with kinetics and characteristics typical of peripheral tissues.

The exogenous administration of different cytokines to the brain parenchyma gives rise to histopathology with distinct characteristics [3], [51]. The injection of IL-1β induces the delayed recruitment of neutrophils but very few monocytes, the neutrophils outnumbering the monocytes by a factor greater than 10 [51]. This is not associated with overt breakdown of the blood–brain barrier (BBB) around the site of injection, although there is acute meningitis and breakdown of the blood–cerebrospinal fluid barrier [3]. In contrast, the injection of TNFα induces a response dominated by mononuclear cells, the mononuclear cells outnumbering the neutrophils by more than 20:1 [51]. Thus, the distinct histopathologies in the brain induced by IL-1β or TNFα suggest that the extensive matrix of cytokine connections, present in the periphery, may not exist in the central nervous system (CNS) parenchyma.

In the present study, we have investigated the acute expression of the pro-inflammatory cytokines IL-1β and TNFα in response to acute pro-inflammatory and excitotoxic challenges in the brain parenchyma of adult rats. The aim of this study was to investigate any differences in cytokine profiles in response to these challenges in the brain parenchyma compared to those of the periphery that may help our understanding of the differences in cellular recruitment.