A role for neutrophil granulocytes as afferent signals in immune-to-brain communication during systemic and localized organ-specific lung inflammation

Abstract

The brain inflammatory response and brain-controlled sickness symptoms, such as fever, involve innate immune system activation through immune-to-brain signaling during systemic inflammatory insults. Neutrophil granulocytes (NG) are a crucial part of the carefully regulated immune system and contribute to the transmission of inflammatory information. Dysregulation of the immune response during an infection can have severe consequences, such as progression to septic inflammation. Indeed, sepsis is characterized by an overwhelming inflammatory response to infection and can be life-threatening or result in long-term impairments due, in part, to poor treatment options. Moreover, neutropenic fever is a severe life threatening condition in immunocompromised patients, such as those that have undergone chemotherapy. Therefore, an anti-inflammatory role for NGs during such inflammatory conditions was hypothesized. This hypothesis is supported by NG-dependent production of specialized pro-resolving mediators (SPM), like resolvin (Rv) E1. For my dissertation project, the role of NGs in two different mouse models of systemic and localized organ-specific lung inflammation was investigated. The first inflammatory model evaluated the effects of neutropenia during severe systemic inflammation. Mice received an intraperitoneal (IP) dose of anti-polymorphonuclear serum (PMN; 1 ml/kg), or normal rabbit serum (NRS) as a control, followed by an IP dose of lipopolysaccharide (LPS; 2.5 mg/kg) or phosphate buffered saline (PBS) 24 h later. The effects on sickness symptoms and peripheral and brain inflammation were investigated. Using a telemetric system, the following physiological parameters were continuously recorded: core body temperature, locomotor activity, and food and water intake. Mice were sacrificed at 4 h or 24 h post-inoculation (p.i.) with LPS and brain and plasma samples were collected. Alterations to peripheral and brain inflammation were assessed by hematological measurements, immunohistochemistry, immunoassays, and RT-qPCR. The second inflammatory model evaluated the effects of genetic omega (ω)-3 polyunsaturated fatty acid (PUFA) enrichment and two different RvE1 receptor (chemerin receptor 23 [CR] or leukotriene B4 receptor [LR]) deficiencies during acute respiratory distress syndrome (ARDS). To investigate the therapeutic potential of SPMs and the role of NGs during ARDS, C57BL/6N mice were bred with (Fat) or without (WT) fat-1 genetic ω-3 PUFA enrichment and crossbred with the RvE1 receptor knock-out (CR KO/LR KO) or unmodified (Norm) mice. Mice were treated with an intra-tracheal (i.t.) instillation of LPS (10 µg/mouse in 50 µl saline) and sacrificed at five different time points: 0 h, 4 h, 24 h, 72 h, or 120 h p.i. Lung, liver, and brain tissues were collected and analyzed by RT-qPCR, multiplex, Western blot, liquid chromatography-tandem mass spectrometry (LC-MS/MS), and immunohistochemistry to assess the peripheral and brain inflammatory response. Interestingly, during severe systemic inflammation, pretreatment with PMN not only reduced circulating levels of NGs but, at higher doses, increased lethality. Moreover, NG recruitment to the brain was attenuated by PMN. All physiological parameters were strongly affected by LPS, but only LPS-induced hypothermia was exacerbated and prolonged in PMN pretreated mice. Indeed, while LPS did increase circulating cytokines, neutropenic mice showed enhanced production of IL-10 as early as 4 h p.i. and by 24 h p.i. plasma levels of CXCL1, CXCL2, IL-10, IL-6, and tumor necrosis factor (TNF) α were all exacerbated in these mice compared to the immunocompetent controls. LPS-induced inflammation in the brain was detectable at 4 h p.i. for inflammatory mediators, enzymes for prostaglandin (PG) E2 synthesis, and markers for activation of inflammatory transcription factors, but these did not persist to 24 h p.i. and were unaffected by PMN pretreatment. However, at 24 h p.i., the mRNA expression of marker proteins for NFκB, STAT3, as well as the enzymes microsomal PGE synthase (mPGES), and cyclooxygenase (COX)2 were generally increased by pretreatment with PMN regardless of LPS treatment. Despite an overall weak PMN effect in the brain, LPS-induced circulating levels of corticosterone were dampened at 4 h but showed a prolonged increase in PMN pretreated compared to NRS pretreated mice. Taken together, these data suggest a peripheral action may be largely responsible for observed alterations in sickness symptoms in neutropenic animals. LPS-induced ARDS resulted in milder inflammation compared to the severe LPS-induced systemic model. The inflammatory response was largely localized to the lung where mediators of inflammation appeared enhanced at 24 h p.i., and alterations by genetic ω-3 PUFA enrichment and RvE1 receptor deficiencies changed over time. These data indicate that both ω-3 PUFA enrichment and RvE1 receptor deficiencies were enough to cause minor alterations in LPS-induced lung inflammation. In contrast, LPS-induced production of inflammatory mediators in the liver was less pronounced. Despite low inflammatory signaling in the periphery (liver), signs of inflammation in the brain were detected in the hypothalamus, similar to the lung with a peak at 24 h p.i. Overall, deficiency in CR had a stronger effect than LR deficiency in the hypothalamus as evidenced by reduced expression of CXCL1 and IL-1β in Fat-CR KO mice compared to Fat-Norm controls. In addition, GFAP expression was increased by CR deficiency regardless of ω-3 PUFA enrichment. Immunohistochemical staining also revealed that NG recruitment to different brain structures were present during inflammation. In combination with the liver’s modest inflammatory response, these effects suggest a minor role of circulating cytokines in lung-to-brain communication. However, NGs trafficking to the brain may contribute to inflammatory signaling from the lung to the brain. Enrichment with ω-3 PUFA and RvE1 receptor deficiencies modulated lipid mediator profiles in the lung and hypothalamus. Such changes in lipids may significantly contribute to lung-brain communication and inflammatory resolution during LPS-induced ARDS. In summary, by analyzing NGs during two different models of inflammation, I found that NGs can be recruited to different brain structures during LPS-induced severe systemic inflammation and ARDS, and potentially participate in immune-to-brain communication. In the case of severe systemic inflammation, neutropenia increases mortality, exacerbates peripheral inflammation, prolongs HPA activation, and worsens inflammatory hypothermia. In this model of inflammation, a peripheral anti-inflammatory action of NGs is most likely contributing to alterations in disease severity. Exaggerated peripheral inflammation during neutropenia could be instrumental in exacerbating sickness symptoms, possibly through interactions with the blood-brain barrier (BBB) or at brain structures with a leaky BBB such as the so-called circumventricular organs. In the milder inflammatory ARDS model, lung-to-brain communication includes low humoral signaling and recruitment of NGs. Both genetic ω-3 PUFA enrichment and RvE1 receptor deficiencies had an effect on the inflammatory response in the lung and hypothalamus. Within the hypothalamus, CR deficiency enhanced GFAP immunoreactivity. The actions of RvE1, as well as ω-3 PUFA enrichment, modulate NG recruitment to the brain most likely through CR. Further investigations into the anti-inflammatory roles of NGs and their ability to participate in immune-to-brain communication are warranted. This will help to gain further insights into underlying mechanisms of the pathophysiology in individuals suffering from neutropenia or when being at risk for brain inflammatory insults as a complication of ARDS.