Bi-directional communication between the periphery and the brain is important for the appropriate response to an immune stimulus . During peripheral infection, pro-inflammatory cytokines are produced in the brain and play a role in adaptive sickness behavior. However, an excessive cytokine response in the brain is associated with prolonged sickness behavior [5–8], cognitive deficits [9–11], and increased anxiety [12, 13]; and the specific role of IL-6 has not been extensively studied. We investigated the capacity of sgp130 to block IL-6 production in microglia and neurons in vitro and explored the effects of sgp130 on LPS-induced sickness behavior in vivo. In vitro, sgp130 attenuated LPS-stimulated IL-6R activation along with IL-6 protein release in both microglial (BV.2) and neuronal (Neuro.2A) cell types. Moreover, in vivo experiments showed that ICV sgp130 facilitated recovery from LPS-induced sickness, and this sgp130-associated recovery was paralleled by reduced IL-6 mRNA and protein levels in the hippocampus.
The present study demonstrates that murine microglia and neurons have the potential to produce IL-6 through both a classic and trans-signaling pathway. In two-color flow cytometry experiments, we showed that BV.2 and Neuro.2A cells expressed both gp130 and IL-6R on their cell surface, though expression differed in each cell type. The findings indicate that classic and trans-signaling are important on both neurons and microglia, though neurons may be more readily regulated through the mechanism of trans-signaling. Previous studies report that the presence of the sIL-6R elicits an exaggerated production of IL-6 protein [32, 34, 35]. Consistent with these reports, we found that pretreatment of sIL-6R led to an IL-6- and LPS-induced increase of IL-6 protein in microglia and neurons. This response is presumably elicited by the ligand and soluble receptor forming a sIL-6R/IL-6 complex. This complex has the ability to bind to the gp130 transmembrane receptor signal transducer and activate intracellular signals that produce IL-6 in any cell type via this trans-signaling mechanism.
LPS binds TLR-4, which we confirmed was present on both microglia and neurons. Upon binding, LPS induces upregulation of the NF-κB transcription factor that binds promoter regions to stimulate the production of IL-6 along with a milieu of other cytokines (e.g. IL-1β, TNF-α, and IL-10) . Soluble gp130 inhibits IL-6 trans-signaling but also regulates IL-6 related cytokines oncostatin M (OSM) and leukemia inhibitory factor (LIF). However, sgp130 has a much lower affinity for OSM and LIF than for the IL-6/sIL-6R complex  and would not be expected to affect either cytokine at the dose used here [19, 37]. Therefore, using sgp130 allowed us to investigate the effects of IL-6 after LPS treatment, while leaving all other cytokines unaffected.
Successful activation of the IL-6R is noted by the dimerization of gp130, resulting in an intracellular cascade that forms recruitment sites for STAT3 in the cytoplasmic region. STAT3 homodimerizes, autophosphorylates, then translocates to the nucleus and binds to enhancer elements of IL-6 to induce gene transcription. Here, STAT3 was upregulated in response to both IL-6 and LPS in BV.2 and Neuro.2A cells and pretreatment with sIL-6R led to an increased IL-6- and LPS-induced STAT3 phosphorylation. However, when pretreated with sgp130, IL-6 and LPS-stimulated BV.2 and Neuro.2A cells displayed a decrease in STAT3 phosphorylation. These data agree with other studies using sgp130 to inhibit IL-6 signaling in peripheral models of inflammation such as arthritis, peritonitis, and colitis [38–40]. To our knowledge, this is the first study to report that pretreatment with sgp130 attenuated LPS-induced IL-6 protein secretion in CNS-derived cells.
LPS activation of the peripheral innate immune system stimulates a robust secretion of inflammatory cytokines through the NF-κB pathway and these cytokines are relayed to the CNS via vagal nerve afferents, and humoral and diffusive pathways [1, 2]. Once in the brain this inflammatory signal is mimicked by innate immune cells [3, 41, 42] and targets neurons which elicit a sickness behavior response that includes general malaise, decreased activity, decreased social interaction, decreased food and water intake (weight loss), and sleep dysregulation [4, 33, 43]. We therefore investigated the effects of ICV sgp130 in vivo and hypothesized that, given the role of IL-6 in neuroinflammatory responses; it would attenuate LPS-induced sickness behavior and IL-6 production. Here we show that sgp130 was effective in facilitating the recovery from LPS-induced social exploratory behavior as early as 8 h in mice. In addition to facilitating recovery from LPS, sgp130 attenuated receptor activation, gene expression, and production of IL-6 in adult mice 8 h after LPS injection. Consistent with previous studies [24, 26, 44, 45], a reduction in brain cytokines did not prevent the initial induction of LPS-induced sickness behavior seen at 2-4 h post-injection, but rather facilitated the recovery from sickness starting at the 8 h time point [32, 34, 35]. In this model, the inability of sgp130 to block the onset of sickness behavior can be attributed to the fact that LPS induces multiple proinflammatory cytokines that have redundant properties and inhibition of a single cytokine is not sufficient to block the initial sickness. It is noteworthy that a study showed that LPS-induced sickness behavior was blocked only if IL-1β, IL-6, and TNF-α were antagonized simultaneously .
This facilitation in recovery from LPS-induced sickness has been observed in various nutritional and pharmacological interventions [5, 47–49] and may be of particular importance when considering conditions where an exaggerated response is elicited during a primed inflammatory state, such as in overexpressing transgenic animals [9, 11], prion disease , and aging [7, 8, 10]. We have previously demonstrated that aged animals display an exaggerated neuroinflammatory and sickness behavior response after activation of the peripheral immune system  and it appears that primed microglia are responsible for this exacerbated phenotype [13, 44, 50]. We and others have shown interventions that are anti-inflammatory are able to ameliorate the exaggerated cytokine response in the brain as well as the mal-adaptive behavioral response that results from peripheral infection [5, 24, 47–49, 51]. Based on the data obtained from this study, it is possible that sgp130 will abrogate the prolonged LPS-induced alterations in sickness behavior, cognition, as well as exaggerated IL-6 levels exhibited in aged mice.