Introduction

Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection [1]. This syndrome frequently affects individuals in intensive care units around the world and presents high rates of mortality and morbidity, with a prevalence of up to 60% in hospitalized patients [2]. In sepsis, the inflammatory response is amplified in such a way that it produces damage to organs other than those affected by the initial infection, and the brain is among those affected, which generates acute and long-term neurological changes [3]. For years, studies have associated the neurological damage in sepsis with an increased permeability of brain barriers caused by the exacerbated release of pro-inflammatory cytokines and peripheral oxidative stress [4], and this would activate immune cells in the brain, such as microglia [5].

Brain signaling can occur both as a response to invasion by microorganisms and in the absence of an infectious agent, as observed in sepsis [6]. Pro-inflammatory cytokines, reactive oxygen/nitrogen species, and patterns associated to tissue damage are among the molecules capable of upregulating brain cells by migrating from periphery through the meninges and brain barriers that protect the CNS; thus, the inflamed brain barriers lose their function of high selectivity, facilitating the influx of toxic molecules into the brain [3]. When passing through brain barriers, such molecules face the brain defense lines, which deposit in the microglia, an inflammatory and precursor cell of macrophages, a “saving” function for neurons. Consequently, microglia are activated and secrete inflammatory mediators, such as cytokines and reactive oxygen species, within the brain parenchyma to combat pathogens and restore homeostasis [7].

However, the mediators released by microglia may play a harmful role, especially in diseases where peripheral inflammation is chronic, sustained, or exacerbated [8]. In this sense, microglia-induced inflammation will activate the cells that compose the brain barriers, thus maintaining their high permeability to peripheral blood and restarting a cell activation cycle [3].

Nevertheless, this cycle does not only involve microglial cells [9]. Possibly, and especially during sepsis, astrocytes are being neglected in terms of their role in protecting from or enhancing neuroinflammation. In a simplistic form, microglia and astrocytes should restore homeostasis; however, here, they lose their role as good guys and become villains. However, these are only cells that are hyper stimulated by the flood of inflammatory mediators from the inflamed peripheral tissues and in situ production by themselves. In this review, we intend to create a new perspective regarding the role of astrocytes during sepsis, encouraging research on this topic.

Astrocytes in the Healthy CNS

Astrocytes represent the most abundant glial cells in the human brain (Verkhratsky & Parpura, 2015). Astrocytes were believed to be a homogeneous population; however, this hypothesis is increasingly being refuted [10]. Like microglia, astrocytes can exhibit different morphologies [11], and based on their morphology and spatial organization, astrocytes are classified into two basic subtypes: protoplasmic and fibrous astrocytes [12]. Protoplasmic astrocytes are found throughout gray matter and show several highly branched and bushy processes, which extend their end-feet to blood vessels and enwrap them to form the glial limiting membrane, which is the outermost wall of the BBB. Fibrous astrocytes, on the other hand, are mainly located within the white matter and have a stellate shape with smooth and long processes [13]. This type of astrocyte expresses high levels of glial fibrillary acidic protein (GFAP) as compared with protoplasmic astrocytes [14].

Astrocytic cells, together with neurons, microglia, pericytes, endothelial cells and the basement membrane, form the neurovascular unit, a structure that involves multicellular relationships to establish a functional coupling between the brain and blood vessels [15]. Astrocytes exhibit different activities to maintain the normal functioning of the CNS, for example, provide almost complete coverage of the cerebral vasculature [16]. Also, with the release of prostaglandins (PGE), nitric oxide (NO), and arachidonic acid (AA), they can control the local blood flow of the CNS [17]. Its role in the metabolism is related to its ability to respond to neuronal activity, being able to increase the rate of glucose uptake, glycolysis, and lactate release in the extracellular space, contributing to neuronal function [18].

Despite the high metabolic activity, astrocytes use at most 15% of the total brain energy in the form of glucose. In fact, the rates of glucose uptake and glycolysis are high; however, as the oxidative phosphorylation rates are low compared to the astrocytic rates, astrocytes serve as a source of lactate production. The lactate is the main substrate for neuronal functioning during cerebral activation, and through the astrocyte-neuron lactate shuttle [19, 20].

In addition to lactate release, astrocytes are part of the tripartite synapses, which are synapses composed of two neurons and one astrocyte. These cells act synergistically with a functional unit for the functioning of plasticity and adequate for the release of neurotransmitters [12, 21].

Astrocytes in Pre-clinical Models of Sepsis

One of the most used animal models of human disease for the study of sepsis is the cecal ligation and perforation model (CLP) [22]. Briefly, a 3-cm midline laparotomy was performed to expose the cecum and adjoining intestine. The cecum was tightly ligated with a 3–0 silk suture in the middle of its length (below the ileocecal valve), perforated once with a 14-gauge needle, squeezed gently to extrude a small amount of feces through the perforation site, and returned to the peritoneal cavity, and the laparotomy was closed with 4–0 silk sutures [23]. This model is well accepted in the international literature for reproducing polymicrobial sepsis and allowing the induction of different degrees of severity, similarly to what is found in humans. Also, this model is effective to study neurological dysfunction during and after sepsis [24].

Astrocytes from CLP animals display an increased gene expression of pro-inflammatory cytokines, such as tumor necrosis factor- α (TNF-α), interleukin (IL)-1β, IL-6, IL-18, monocyte chemoattractant protein-1 (MCP-1), and cyclooxygenase-2 (COX-2), with a reduction in IL-10, accompanied by augmented levels of Toll-like receptor (TLR)-2 mRNA expression but no changes either in TLR4 or in vascular endothelial growth factor (VEGF) gene expression [25]. These results are like those found in rat brain tissue homogenate after CLP [26]. This cytokine profile has been closely related to the action of microglial cells [27], but interestingly, a study provided evidence that mediators released by peripheral blood mononuclear cells (PBMC) directly promote astrocyte reactivity during sepsis, independently of microglia [28], corroborating our previous finding that sepsis causes astrocytic activation [26].

The Table 1 summarizes the main findings concerning CLP-induced sepsis effects on astrocytes. The studies point the occurrence of astrocytic activation as early as 4 h after polymicrobial sepsis [29], which may persist for up to 15 days after sepsis [30]. The hippocampus was the most prevalent brain structure evaluated in the studies.

Table 1 The effects of polymicrobial sepsis induced by CLP on astrocytic activation in animal models
Full size table

The animal model of sepsis by CLP generates endotoxemia in animals, especially due to the presence of gram-negative bacteria. Therefore, the administration of lipopolysaccharide (LPS) alone can mimic the inflammatory response that occurs in sepsis in animal models of human diseases. The effects of LPS on astrocytic activation in animal models are shown in Table 2. The systemic effects of LPS induce important changes in the brain of animals submitted to sepsis, and these studies demonstrate a more mechanistic evaluation involving the astrocytes, although they also show astrocytic activation and release of pro-inflammatory cytokines provoked by LPS injection.

Table 2 The effects of LPS injection on astrocytic activation in animal models
Full size table

However, most information about the role of astrocytes comes from studies in cell culture. However, in sepsis, these profiles are still little explored; Table 3 presents the studies carried out with astrocyte culture stimulated essentially with LPS. In addition to the astrocytic activation being evident, there is evidence of alteration in inflammatory mediators, which supports the hypothesis of a greater influence of these cells on the pathophysiology of sepsis. Recent, study has shown that neuroinflammation or ischemia induced two different types of reactive astrocytes, referred to as “A1” and “A2” [43]. It was shown that the A1 phenotype positively regulates different classical complement cascade genes and complement 3 (C3), which is harmful to neurons and oligodendrocytes. On the other hand, the A2 phenotype induced by ischemia positively regulates many neurotrophic factors that promote the survival and growth of neurons [44]. In fact, melatonin was able to reduce the number of A1 astrocytes and increase the number of A2 astrocytes in the periventricular white matter of neonatal septic rats [45].

Table 3 The effects of LPS on astrocyte cell culture
Full size table

Although there is a lot of evidence about the activation of astrocytic cells after sepsis or LPS stimulus, however, there is no direct evidence of the interaction between these cells’ with the BBB. The role of BBB dysfunction after sepsis has been extensively studied, and its impairment is known to play an important role in acute neuroinflammation and long-term damage after sepsis [3]. However, in some other publications, the authors strongly suggest that astrocytes release several substances that regulate the function of the BBB since it is formed [56, 57]. In this sense, probably, the changes in astrocytic profiles could influence mostly the BBB function after sepsis, leading to increase the neuroinflammation and the neuronal damage e after this the cognitive decline.

Astrocytes in the Human Septic Brain

Data about the astrocytes activation in humans are scarce; however, evidence suggests that, as occurs with microglia, astrocytes also change from a resting to an activated state when stimulated [58]. Human brain astrocytes are more diverse, more complex, and in greater numbers when compared to the rodent brain [59, 60].

In the post-mortem analysis of tissue from the right frontal lobe of 3 sepsis cases, an elevation of the GFAP marker was identified, thus indicating high astrocytic activity in the region when compared to controls [61]. Other studies with pediatric, adult, and elderly septic patients show in common an elevation of the s-100β marker at the expense not only of microglial activation, but also of astrocytic activation. This activation was associated with longer delirium duration, higher delirium severity, and in-hospital mortality [12, 62, 63].

Conclusion

In summary, there is robust preclinical evidence of astrocytic activation after sepsis and few clinical studies. The inflammatory stimulus generated by polymicrobial sepsis (CLP model) or by the LPS stimulus can generate systemic responses that influence the functioning of astrocytes in the brain (Fig. 1). However, little is known about the effects of this activation on the pathophysiology of sepsis. In view of the data presented here, the emerging need to establish the role of astrocytes in the pathophysiology of neurological dysfunction during sepsis becomes evident. Some questions still need scientific evidence, for example, how homeostatic functions of astrocytes are affected in sepsis, how occurs the astrocyte activation after BBB activation, and how astrocytes interact with the BBB after this. These and other questions may explain how astroglial activation can contribute to the long-term consequences of sepsis.

Fig. 1
figure 1

Role of astrocytes in sepsis development. During sepsis, pathogens, inflammatory and oxidative mediators are released in the blood. These mediators interact with the cells of blood brain barrier, increasing your selective permeability (1). Lipopolysaccharide (LPS) and other molecules entry in the brain parenchyma (2), lead to activation of glial cells, such as astrocytes and microglia. Activated astrocytes can activate microglia cells (3), besides release reactive oxygen species (4) and cytokines (5). These events culminate in neuronal damage (6). On the other hand, stressed neurons can also activate microglia (7) and increase de neuroinflammatory response in septic brain. There is also evidence of astrogliosis during sepsis (8). Thus, astrocytes may play a much more important role in sepsis than has been attributed to them

Full size image