Evolutionary Aspects of Infections: Inflammation and Sickness Behaviors

Abstract

Sickness behavior was conceptualized initially as the behavioral counterpart of the fever response to infectious pathogens. It helps to raise body temperature to its higher setpoint and to maintain it at this new level and it has the additional benefit of enabling a weakened organism to protect itself from other dangers. The discovery of the behavioral effects of proinflammatory cytokines produced by activated immune cells provided a cellular and molecular basis to this phenomenon. The administration of cytokines or cytokine inducers like lipopolysaccharide to healthy rodents allowed to reveal the similarities and differences between inflammation-induced sickness behavior and the fever response. It also led to the understanding of how the inflammatory response that is triggered at the periphery can propagate into the brain and induce the behavioral manifestations of sickness. At the behavioral level, the demonstration that sickness behavior is the expression of a motivational state that reorganizes perception and action in face of a microbial pathogen just like fear in face of a predator appeared at first glance to strengthen the adaptive value of this behavior. However, all aspects of sickness behavior are not always favorable for the organism. This is the case for anorexia that is beneficial in the context of bacterial infection but detrimental in the context of viral infection. In addition, studies of sickness behavior in natural conditions revealed that like any other defensive behavior, sickness behavior requires trade-offs between its survival benefits for the sick individual and the costs incurred especially in the context of gregarious groups. Thanks to these studies, evidence is emerging that sickness behavior is much more variable in its expression than initially thought, and that part of this variability depends not only on the pathogen and the social context in which the infection develops but also on individual factors including species, sex, age, nutrition, and physiological status.

I. Introduction

While I was studying veterinary pathology at the School of Veterinary Medicine in Toulouse, France, in the mid 1960s, I was struck by the archetypal description of what sick animals look like. The image was always the same whatever the animal species. At the beginning of their infectious episode, sick animals are febrile and show lethargy, depression, anorexia, and reduction in grooming. They isolate themselves from their conspecifics and have no appetite even for the type of food they normally favor. Of course, the experienced veterinarians whom we all aspired to become had to go beyond these non-specific symptoms of infection and search for those clinical signs that are specific of the infectious disease itself. The behavior of sick animals was not interesting by itself as our ultimate objective was to identify the disease in order to treat it in the best possible way. Still, what was intriguing for me at the time was that the conceptualization of this non-specific behavioral pattern of sickness was not without analogy with the discourse on the non-specific activation of the pituitary-adrenal axis that is characteristic of the stress response as defined by Hans Selye. This idea of sickness behavior as a stereotyped response to infection remained with me for a long time but without any real willingness to address it on my part. I had enough on my plate with my research work on stress and well-being.

The game changing event occurred in the late 1980s when molecular biologists were able to characterize the inflammatory mediators that are released by activated innate immune cells and mediate the development of infection-induced fever and activation of the hypothalamic-pituitary-adrenal response. After getting enough recombinant human interleukin-1β IL-1β) from Glaxo I was finally able to demonstrate for the first time that this proinflammatory cytokine mediates the development of sickness behavior as well [1–3]. Before their molecular structure was elucidated IL-1β and other proinflammatory cytokines were known as endogenous pyrogens, i.e., the intermediate factors that the host needs to produce in order to develop a fever response to an infectious agent. Sickness behavior was first conceptualized in the context of the fever response [4]. Fever is the result of an increase in the set point for the regulation of body temperature, meaning that the febrile individual must increase its heat production and at the same time drastically decrease its heat losses. One cannot develop a fever if one has to engage in the expensive activity of foraging or searching for a sexual partner. Increased sleepiness and lethargy suppress all activities directed toward the environment. Huddling minimizes body area exposed to ambient air while reduced grooming minimizes heat losses. In other words, it made sense to claim that sickness behavior was the necessary behavioral counterpart of the energetic requirements of the fever response to infection.

The objective of this chapter is to present the mechanisms of sickness behavior, show its universality across animal species equipped with an innate immune system, and discuss the validity of the claim that because of its association with the fever response it is always adaptive for the host.

2. Neuroimmune mechanisms of sickness behavior

As previously mentioned, sickness behavior is characterized by lethargy, sleepiness, anorexia, and social isolation. In the terms used by Hart, “animals that are acutely ill with systemic protozoan, bacterial or viral infections are typically described as depressed and lethargic with little interest in eating food and drinking water. A little later in the course of a disease they commonly show signs of dehydration along with indications that they have lost interest in grooming since they develop rough hair coats. These behavioral signs generally accompany a fever response and, together with the occurrence of fever, are recognized by animal handlers and veterinarians as signs that an animal is sick or is becoming sick with an infectious disease”. However, this does not mean that all animals that develop these signs are always afflicted with an infectious disease as similar behavioral alterations can be observed in several other conditions, such as poisoning or dehydration. Infection-induced sickness behavior can be identified more specifically by its dependency on the mechanisms that mediate the recognition of infectious agents by innate immune cells, the production of proinflammatory cytokines at the site of infection, and the propagation of this immune message to the brain.

Innate immune cells recognize pathogen-associated molecular patterns (PAMPs) which correspond to specific molecular motives of infectious agents. This is the case for instance for lipopolysaccharide (LPS) that is the major component of the outer layer of the membrane of Gram-negative bacteria. PAMPs are recognized by pattern specific receptors that are expressed by innate immune cells, such as monocytes, macrophages, neutrophils, dendritic cells, and epithelial cells [5]. Toll-like receptors (TLRs) are a well known example of pattern recognition receptors. They are composed of an extracellular domain, a transmembrane domain, and an intracellular effector domain that recruits adaptor molecules to activate downstream signaling pathways. For example, upon binding LPS TLR4 recruits the adaptor myeloid differentiation factor (MyD88) that ultimately signals via the transcription factor nuclear factor-kappa B. This leads to the production of proinflammatory cytokines such as IL-1, tumor necrosis factor, and IL-6. Not all pattern recognition receptors are located at the cell membrane. Some receptors are located intracellularly. They are important for the recognition of molecular motives of infectious agents that invade cells. This is the case for nucleotide oligomerization domain (NOD)-like receptors and retinoic acid-inducible gene-I (RIG-I)-like receptors. NOD1 for instance recognizes a component of the cell wall of Gram-negative bacteria whereas NOD2 recognizes single-stranded ribonucleic acid of viruses. Upon binding to their ligands, NOD-like receptors are recruited into the plasma membrane and endosomal membrane where they initiate signal transduction. RIG-I-like receptors recognized double-stranded RNAs and their activation result in the increased expression of type I interferons.

Activation of pattern recognition receptors leads to the production of cytokines via multiple signaling pathways. The NFκB signaling pathway mediates the innate immune response and its interaction with the adaptive immune response. It promotes the transcription of most inflammatory genes other than interferons for which IRF-3 downstream of TLR3, the TLR4-TIR domain and RIG-I are the key elements. Mitogen-activated protein kinase (MAPK) signaling relays the effects of inflammatory cytokines on their cellular targets. The inflammasome is a multi-protein complex that is assembled in the cytoplasm by pattern recognition receptors and results in the activation of caspase-1 which is necessary for maturation of cytokines of the IL-1 family. Most inflammatory cytokines act as autocrine or paracrine communication signals in the local environment in which they are produced. IL-6 is an exception. It is released into the general circulation and acts on hepatocytes to promote the production and release of acute phase proteins such as C-reactive protein.

The way cytokines ultimately act in the brain to induce sickness behavior has been studied mainly in rats, guinea pigs, and mice, and has been the subject of much debate [6]. This debate was fueled initially by the use of the intravenous route to inject endogenous pyrogens. The predominant view at the time was that macrophages release endogenous pyrogens into the general circulation. Endogenous pyrogens cannot enter the blood-brain barrier. Instead, they activate macrophage-like cells in circumventricular areas, those brain regions that lack a fully formed blood-brain barrier. There, they induce the formation of secondary messengers in the form of prostaglandins E2 that can diffuse freely through the blood-brain barrier to reach neurons of the thermoregulatory center of the hypothalamus. However, the discovery that at least part of the effects of cytokines on the brain are relayed by sensory nerves conflicted with this theory [7–9]. Cytokines activate directly or indirectly sensory neurons that innervate the site of inflammation. This neural form of information is then transmitted to the brain via the primary and secondary projections of these neurons [10]. As a typical example, proinflammatory cytokines released in the abdominal cavity by LPS injected via the intraperitoneal route activate the sensory branch of the vagus nerves that project to the nucleus tractus solitarius at the level of the brain stem. This neural activation then propagates to other brain areas such as the parabrachial nucleus, paraventricular and supraoptic nuclei of the hypothalamus, central amygdala, bed nucleus of the stria terminalis, and periacqueductal grey. This neural pathway intersects with a humoral pathway that is dependent on the passage of pathogen-associated molecular patterns in the general circulation. By recruiting macrophages in circumventricular areas this humoral pathway is at the origin of a slow wave of production of proinflammatory cytokines that progress into the brain parenchyma by active diffusion, recruiting other macrophage-like cells “en passant”. This secondary wave of cytokines originating from circumventricular areas would be responsible ultimately for the production of proinflammatory cytokines by microglial cells which are the resident macrophages in the brain parenchyma. It has been proposed that the neural pathway traces the path for the effects of these locally produced cytokines by sensitizing the brain areas to their activity [6]. However, this hypothesis has not yet been tested. Another pathway is represented by the recruitment of inflammatory monocytes to the brain vasculature by microglia [11]. This would allow the trafficking of these peripheral innate immune cells into the parenchyma of those brain regions in which the blood-brain barrier is compromised. This pathways appears to be activated in chronic stress conditions such as the repeated social defeat paradigm [12].

We do not yet fully understand what happens when the immune message propagates to the brain. Several neuronal effects of cytokines are indirect and relayed by the synthesis and release of inflammatory mediators such as prostaglandin of the E2 series, radical oxygens species or nitric oxide. The main cellular source of prostaglandins is represented by endothelial cells and perivascular cells along small venules. Prostaglandins act on various prostaglandin receptors that are distributed throughout the brain in neuronal circuits that have been found to mediate the effects of inflammatory cytokines on fever, sleep, anorexia, activation of the hypothalamic-pituitary-adrenal axis, and hyperalgesia [13]. However, other mechanisms could also be at play as IL-1β, when it is expressed in the brain, can act directly on neurons by increasing or decreasing their excitability depending on the amount of IL-1β present [14].

Microglial cells produce inflammatory cytokines in the brain either in response to inflammatory stimuli taking place in the brain or in response to peripheral inflammation relayed to the brain by the mechanisms described above. The role of microglia in the development of sickness behavior has been studied mainly using pharmacological tools. Administration of minocycline, a second generation antibiotic that acts as a down regulator of microglia activation, attenuated the effects of LPS administered peripherally on mood [15]. In the same manner, minocycline treatment of mice exposed to repeated social defeat, a stressor procedure that induces microglia activation and monocyte recruitment in the brain, abrogated these effects together with the cognitive deficits presented by mice submitted to the social stressor [16]. Interestingly, this intervention did not prevent the persistence of social avoidance behavior that was apparent in stressed mice. However, a more radical intervention consisting of depleting brain microglia by various modalities of intervention not only failed to abrogate inflammation-sickness in mice but actually enhanced its expression probably by unmasking an inhibitory role of microglia on the immune activity of astrocytes [17]. More recently, specific activation of microglia in the mouse dorsal striatum by chemogenetic tools was shown to induce anhedonia and aversion. These effects were mediated by IL-6 signaling in microglia and Cox1-dependent production of prostaglandins acting on EP1/2 receptors In striatal medium spiny neurons [18]. Conversely, chemogenetic inhibition of microglia abrogated the development of aversion in LPS-treated mice.

The emphasis on the neuroimmune mechanisms of sickness behavior has led researchers to neglect the fact that this behavior is by necessity associated with important alterations in energy metabolism that are not only dependent on the requirements of the fever response as initially thought but also by those of activated immune cells. The proliferation of immune cells and production of proinflammatory cytokines necessitate a reprogramming of their cellular metabolism from oxidative phosphorylation to aerobic glycolysis. The molecular mechanisms of this metabolic reprogramming have been elucidated thanks to advances in immunometabolism [19]. Whatever the details of these mechanisms it is important to remember that aerobic glycolysis generates only 4 molecules of adenosine triphosphate, the currency for energy metabolism in the organism, for one molecule of glucose instead of 36 for oxidative phosphorylation. The high increase in energy metabolism at the site of inflammation can only be sustained if other parts of the organism including the brain and skeletal muscles consume less energy. This means that sickness behavior is probably dictated at least in part by energy repartition considerations [20] or by the delicate balance between defense and tolerance metabolic programs within the organism [21]. We will come back to this notion in the section discussing the adaptive value of sickness behavior.

3. Evolutionary aspects of sickness behavior

The exact mechanisms by which proinflammatory cytokines act in the brain are likely to depend on the species. What is important for the identification of sickness behavior as a component of the defense response of the organism to an infection is that this behavior should not only occur during the course of an infection, but its development should be dependent on the production and release of inflammatory mediators by innate immune cells. As mentioned in the previous section, the mechanisms of infection-induced sickness behavior have been mainly studied in laboratory rodents. However, infection-induced sickness behavior is also present in other species including insects, fish, reptiles, and worms [22]. As a typical example, Kirsten and colleagues recently demonstrated that zebrafish inoculated with formalin-inactivated Aeromonas hydrophila bacterin developed a reduction in their locomotor activity, social preference, and exploratory behavior towards a novel object, and these changes in behavior were associated with an upregulation of the expression of proinflammatory cytokines in their brains [23]. This allowed these authors to conclude that the divergence in cytokine profile they observed between individual zebrafishes differing in their response to novelty and to social stimuli [24] was probably driven by neuroimmune interactions. When the microbial agent invades the brain like it was described in zebrafish infected with the tilapia lake virus, the behavioral signs of infection were even more severe with signs of neuropathology developing over a background of sickness behavior [25].

The innate immune response to microbial pathogens in drosophila involves two pathways. The Toll pathway regulates production of anti-microbial peptides against fungi and Gram-positive bacteria. The immune deficiency (IMD) pathway responds to Gram-negative bacteria [26] by also producing antimicrobial peptides. The ability of anti-microbial peptides to act on the nervous system has mainly been studied in the context of sleep. Increased sleepiness in drosophila is a well characterized response to microbial infection and to other forms of cellular stress. In Caenorhabditis elegans, anti-microbial peptides act as somnogens, signaling across tissues to promote sleep by activating sleep neurons in response to peripheral tissue injury [27]. In Drosophila melanogaster, the antimicrobial peptide nemuri can function both at distance when produced in damaged peripheral tissues and locally as it is expressed also in the brain where its enhanced production following sleep deprivation promotes restorative sleep [28]. In general, mechanisms regulating sleep during sickness are partially distinct from those regulating healthy sleep as demonstrated by elegant experiments in C. elegans and drosophila. Cellular stress induced by infection or other environmental insults is associated with the release of epidermal growth factor (EGF) or EGF-like peptides that signal to central neuroendocrine cells. These neurons release RFamide peptides that induce anorexia, movement quiescence and elevated arousal threshold [29]. Like in vertebrates, cooperation between the immune system and the nervous system in invertebrates involves a variety of mechanisms that allow animals to adapt their behavior to the presence of pathogens that coexist with them [30].

Although the behavioral effects of proinflammatory cytokines in non-human primates has been less well studied than in laboratory rodents, there is evidence that they are able to develop the whole range of sickness behaviors observed in other animal species. For instance, intravenous injection of IL-1α in juvenile rhesus monkeys rapidly induced sleep-like inactivity and decreased their behavioral and vocal responses to broadcasted calls from conspecifics [31]. Monkeys injected with a high dose of IL-1α showed signs of agonistic behavior when challenged by a human experimenter [32]. Similar signs of inactivity and increased sleepiness were observed in response to IL-1β in monkey pairs with evidence of enhanced huddling behavior [33]. This tendency to increased social contact was observed also in monkeys injected with LPS and is in contrast with the social withdrawal seen in rodents in response to the same treatment [34]. All these behavioral changes were associated with increased circulating levels of IL-6 although its role in the development of sickness behavior was not specifically addressed.

As studies of sickness behavior in humans are very disparate in terms of research design, modality of immune stimulation, and selection of the endpoint under study, a meta-analysis of the human sickness behavior was conducted recently [35]. Depressed affect and fatigue were the most commonly reported symptoms, and they showed an association with IL-6 and IL-1β. Focusing on LPS-induced sickness behavior, Lasselin and colleagues confirmed that reduction in food intake and alterations in sleep patterns are part of the picture [36]. Other objective behavioral changes include increased frequency of moans and sighs, increased yawning, reduced walking speed, and changes in the willingness to expend effort to get a reward. Self-report assessments provide clear evidence for fatigue, reduction in appetite, and reduced social interest, together with an increase in negative mood as well as in state anxiety. All these effects develop within 2–3 hours after intravenous injection of LPS and are over by 6 hours. Experiments specifically designed to address the impact of LPS on social experiences revealed that inflammation increases sensitivity to negative, threatening social experiences, and in contrast increases sensitivity to positive, socially rewarding experiences [37].

The behavioral effects of immune stimulation targeting TLR3 or intracellular DNA sensors and mimicking a viral infection have been less studied than those targeting TLR4 and mimicking a Gram-negative bacteria infection. Acute peripheral administration of polyriboinosinic:polyribocytidlic acid (poly I:C), a synthetic double-stranded RNA that mimics a viral infection, induces a dose-dependent typical sickness behavior in mice and rats, with decreased locomotor activity, feeding and burrowing, and hyperthermia or hypothermia depending on the dose [38]. Although there was no evidence of tolerance when the treatment was repeated, most experiments on poly I:C make use of only acute injections. The study of the behavioral effects of poly I:C are complicated by the existence of different molecular weight forms of this compound which do not have the same ability to activate the type I interferon response [39]. In addition, there is evidence that the reduction of voluntary wheel running in response to poly I:C, a model of inflammation-induced fatigue, is independent of its ability to induce IFNβ [40]. The anorexic response to poly I:C was blocked by genetic deletion of TLR3 or TRIF when the compound was injected at the periphery but not when it was injected into the lateral ventricle of the brain [41]. The central effect of poly I:C was blocked by genetic deletion of MyD88, suggesting that the mechanisms of sickness induced by a neurotropic virus could differ from those of sickness induced by a non-neurotropic virus.

Interferons have been studied mainly in the context of anxiety disorders and depression. The behavioral effects of these cytokines will not be detailed in the present chapter. Casual observations confirm that acute administration of interferons induces symptoms of sickness although they have not been studied systematically. Chronic administration of interferon-α, induced anxiety-like behavior and decreased environmental exploration in rhesus monkeys [42]. Locomotor activity was decreased only in dominant monkeys whereas subordinate monkeys exhibited increased locomotor activity. Three monkeys out of eight developed persistent huddling behavior which was interpreted as a sign of depression-like behavior.

The effects of inflammation on learning and memory need to be mentioned despite the fact they are not really part of the description of sickness behavior. Although it is usually assumed that proinflammatory cytokines negatively interfere with learning and memory, there is a need for some nuance. Administration of LPS to rats for instance abrogated acquisition of a fear response conditioned to the context but had no effect on the fear response to the distinctive cues paired with electric shocks [43]. Administration of LPS or IL-1β to rodents does not induce reliable effects on spatial reference memory measured in the Morris Water Maze and it is difficult to separate the observed effects from the impact of sickness on performance [44]. Furthermore, the conditioned taste avoidance induced by LPS or IL-1β attests that inflammation does not prevent the association of taste cues with sickness [3, 45].

4. The adaptive value of sickness behavior

As mentioned in the introduction, sickness behavior is considered traditionally to be a highly organized defense strategy that is adopted by a sick individual to facilitate defense against an invading pathogen [4]. This notion implies that sickness behavior is part of a motivational system that reorganizes the host priorities in face of a microbial infection in order to maximize its fitness just like fear in face of a physical danger [46]. However, this does not guarantee that sickness behavior is always adaptive for the host. In the same way that the increased vigilance of a fearful individual toward all possible danger objects compromises its ability to benefit from other opportunities offered by its environment, the survival benefits of sickness behavior in terms of conservation of metabolic energy for the immune system comes with important costs represented by increased risk of predation, reduction in social engagement and reproductive opportunities, and diminished territorial defense [47]. Detailed observations of acute infections in wild vervet monkeys showed that sick individuals were actually twice as likely to receive aggression from conspecifics and six times more likely to become injured than healthy individuals [48]. In addition, fevers did not influence the time spent socializing with conspecifics, suggesting that social isolation is not a constant feature of sickness behavior and, by extension, that sickness behavior does not necessarily decrease the risk of disease transmission within the group. These original findings indicate clearly that there is a need to assess more in depth the trade-off between the survival benefits of sickness behavior and the costs incurred especially in the context of gregarious groups. This is the object of a specialized discipline known as ecoimmunology [49].

Some of the issues revolving around the adaptive value of sickness behavior are the same as those that have been addressed already for the fever response. It is clear that there is no simple answer to the question of the usefulness of treating fever, sickness behavior, and the underlying inflammatory response by antipyretics, in the same way that it not possible to prove that fever and sickness behavior are beneficial under all circumstances [50, 51]. The matter is made even more complicated by the fact that it is not possible to generalize from one type of infection to another one. For instance, there is evidence that one of the most common signs of sickness behavior, anorexia, is protective in the case of the bacterial sepsis that develops in mice infected with Listeria monocytogenes while nutritional supplementation is detrimental [52]. Similar findings have been reported for invertebrates as well [53]. However, nutritional supplementation protects against mortality from influenza, a viral infection, whereas blocking glucose utilization is lethal in this condition [52]. This observation has been interpreted to suggest that distinct inflammatory responses are coupled with specific metabolic programs that support either resistance resulting in elimination of the pathogen or tolerance consisting of living with the pathogen and minimizing the tissue damage it can cause. In the case of a bacterial infection, ketogenesis is the predominant metabolic program while glucose utilization predominates in the case of a viral infection. Not surprisingly, resistance to pathogen infection via fever and sickness behavior has been the most commonly studied response in immunopathology whereas much less is known about tolerance and the way it interacts with resistance to ultimately determine health [54]. Whether tolerance is associated with lack of sickness behavior or a form of sickness behavior that promotes tolerance has not yet been addressed.

Social distancing is another feature of sickness behavior that has received some attention [55]. Sick individuals isolate from the social group either passively because of their physical weakness or the avoidance response they trigger in healthy members of the group, or actively, by isolating themselves from other members of the group. This behavior appears a priori to have survival value not for the sick individuals themselves but for the group in which they belong as it minimizes the risk of dissemination of the pathogen. However, the value of social isolation as a strategy to fight infection cannot be generalized. We have seen earlier that feverish vervet monkeys do not isolate from other members of the group in natural settings. Social immunity in insects – the fact that insects living in group develop the capability of mounting collective anti-pathogen defenses [56] - work against social isolation as the best strategy to fight infection. Of note, social isolation of the sick individual by its conspecifics requires the ability to recognize sickness cues emanating from the infected individual and their use to trigger avoidance responses, a competence that is probably not shared to the same extent across different species. In addition, the recognition of sickness in infected individuals should not always trigger an avoidance response by the rest of the group, otherwise there would be no possibility of caregiving [57].

Another important question in terms of adaptation is for whom sickness behavior is advantageous. So far, we have focused on the adaptive value of sickness behavior for the host or for the group in which it belongs. However, it is easy to imagine that the sickness response can be exploited by the pathogen to its advantage. If the sickness induced by the infectious pathogen spares locomotor activity, its dissemination will be favored as the host continues to be active instead of developing lethargy and socially isolating itself. This type of strategy appears to be exploited by Salmonella typhimurium. This Gram-negative bacterium does not cause anorexia as it has the ability to inhibit inflammasome activation and IL-1β maturation in the small intestine, therefore limiting transmission of the immune message from the gut to the brain [58]. This allows the infected host to still engage in foraging, which increases the risk of transmission of the pathogen to new hosts thanks to the fecal shedding of the infectious agent.

5/. Conclusion

What we know about sickness behavior has been acquired mainly through very well controlled laboratory experiments consisting of exposing genetically homogenous individuals not to microbial pathogens of which the infectious nature and virulence can vary depending on the stock but to specific pathogen-associated molecular patterns. Most of the work has been done on lipopolysaccharide that targets TLR4 and partly TLR2. This has allowed us to better understand the mechanisms of sickness behavior. We now view this form of behavior as an adaptive response to the sensing of inflammatory mediators by the brain. Its postulated function is to serve to maximize the efficacy of the inflammatory response by favoring reallocation of metabolic energy to the immune system. However, evidence is emerging that sickness behavior is much more variable in its expression than initially thought, and that part of this variability depends not only on the pathogen and the social context in which the infection develops but also on individual factors including species, sex, age, nutrition, and physiological status. In addition, although the mechanisms that lead to the development of sickness behavior have been well characterized at least in the context of TLR4 signaling, much less is known about the mechanisms that are responsible for its dissipation once the pathogen has been cleared or the organism has shifted its strategy from defense to tolerance. The persistence of some aspects of sickness behavior once the infection has been cleared at least in appearance is still an important issue in the context of pathologies such as the post-sepsis syndrome or its likely family member, long Covid.