Neuroimmune modulation of pain across the developmental spectrum

Abstract

Today’s treatment for chronic pain is inadequate, and novel targets need to be identified. This requires a better understanding of the mechanisms involved in pain sensitization and chronification. In this review, we discuss how peripheral inflammation, as occurs during an infection, modulates the central pain system. In rodents, neonatal inflammation leads to increased pain sensitivity in adulthood by priming immune components both peripherally and centrally. The excitability of neurons in the spinal cord is also altered by neonatal inflammation and may add to pain sensitization later in life. In adult humans, inflammation modulates pain sensitivity as well, partly by affecting the activity in brain areas that process and regulate pain signals. Low-grade inflammation is common in clinical populations both peripherally and centrally, and priming of the immune system has also been suggested in some pain populations. The nociceptive and immune systems are primed by infections and inflammation. The early life programming of nociceptive responses following exposure to infections or inflammation will define individual differences in adult pain perception. Immune-to-brain mechanisms and neuroimmune pathway need further investigation as they may serve both as predictors and therapeutic targets in chronic pain.

Introduction

Inadequate treatment of pain represents a public health crisis worldwide [1–3]. In the USA alone, roughly 100 million individuals are living with ongoing pain and an estimated 20 million of these patients live with high-impact chronic pain resulting in substantially restricted work, social, and self-care activities [4,5]. The financial toll of chronic pain in the US alone exceeds half a trillion dollars per year. We need to better understand the neurobiological mechanisms underlying pain (and its chronification after injury) and translate that knowledge into safe and effective therapies.

A target for future pain treatment may in fact be the immune system. While sensitization of peripheral nerves during peripheral inflammation has been studied extensively, less attention has been directed toward the central effects of peripheral inflammation. In this review, we highlight how peripheral immune activation modulates both the development and function of central pain pathways. Researchers have recently demonstrated that peripheral immune activation has profound effects on the central nervous system [6,7••]. Immune-to-brain interactions have been explored most intensely in depression [8], but appear relevant for the development of chronic pain as well [9]. The individual differences in the nervous system and the immune system early in life will dictate how this neuroimmune communication manifests in the adult individual.

The developmental scheme of the nervous system is to transition from an overproduction of neurons and synapses, to a pruning of required pathways as guided by sensory experience [10,11]. In contrast, the immune system starts with rudimentary components that are enhanced and developed through interactions with microbes in the environment [12,13]. The nervous system is plastic throughout life, but particularly during early childhood. Stress affects the developing nervous system [14,15], with implications for both pain and mood disorders later in life. Likewise, early life painful experiences affect nociception and pain behaviors throughout life [16,17]. For instance, neonatal exposure to repeated hindpaw needle pricks in rats, an animal model that mimics the invasive procedures experienced by preterm infants admitted to the neonatal intensive care unit [18], has been shown to enhance sensitivity to noxious stimuli in the adult. Infections, and the inflammatory cascade stemming from them, are also strong modulators of the early pain system. Here, we discuss how peripheral inflammatory activity, such as during an infection or during low-grade chronic inflammation, modulates the pain system from birth to adulthood. We describe early life modulations in rodents and discuss translational aspects in human experimental models and clinical populations.

Programming of pain sensitivity following neonatal inflammation

Neonatal exposure to lipopolysaccharide (LPS) causes a functional bidirectional interaction between the immune and nervous systems, both at the peripheral and central levels, in an attempt to maintain tissue integrity in the periphery and synapse homeostasis in the CNS [11,19]. Upon exposure to LPS, in both young and adult rodents, immune cells such as neutrophils, monocytes, macrophages, and mast cells are activated and release a plethora of proinflammatory cytokines. These include, but are not limited to, interleukin (IL)-1β, tumor necrosis factor (TNF)-α, and IL-6, and inflammagens such as histamine, released as a consequence of degranulating mast cells. These cellular reactions ultimately create an inflammatory soup in the blood or in the local tissue that contributes to enhanced pain responses during the active phase of the inflammation [20]. The peripheral inflammatory soup not only affects the peripheral nerve fibers, but also induces central neuroimmune activation via humoral and neuronal pathways. This leads to the production of cytokines in the spinal cord as well as key brain regions involved in pain modulation, such as the prefrontal cortex, hippocampus or amygdala (for thorough reviews of sickness behavior see e.g. Ref. [21]). Neonatal exposure to LPS has also been reported to increase pain behaviors, which is maintained in neonatal, preadolescent and adult rats [22,23,24•,25], long after the inflammation is resolved, which suggests structural changes of the pain system after infection.

Peripheral inflammation leads to long-term peripheral and central immune changes that influence pain sensitivity

The long-term programming of pain sensitivity after neonatal LPS exposure appears to occur through changes at the peripheral and central immune response. At the peripheral level, exposure to LPS (50 μg/kg) at postnatal days (PND) 3 and 5 in rats elevated plasma IL-1b and produced mast cell degranulation at PND 22 compared to control saline-injected animals, indicating sensitization of the peripheral inflammatory response following bacterial infection. Centrally, astrocytes are key modulator of inflammation and regulate cytokine levels in the hippocampus following inflammation [26]. Intraperitoneal injection of LPS (100 μg/kg) in rats at PND 21 elevated proinflammatory cytokine mRNA levels (IL-1β, TNFα, IL-6) and astrocyte activity (i.e. GFAP mRNA in the lumbar spinal cord [23]. Supraspinally, neonatal LPS exposure in rats enhanced FOS immunoreactivity in the central nucleus of the amygdala (CeA) at PND 14 [24•] and reduced FOS activation in the periaqueductal gray (PAG) [27]. This indicates that neonatal inflammation sensitizes the neural circuit modulating pain, by enhancing activity in key brain areas that modulate the emotional component of pain and concomitantly decreasing activity in the descending inhibitory system, resulting in exaggerated pain responses. Moreover, neonatal LPS enhances hippocampal IL-1β following formalin injection or stress in adult animals [25,28]. Therefore, neonatal inflammation can not only sensitize the neural circuit modulating pain, but also increase the inflammatory reactivity in the central nervous system, leading to a ‘primed brain state’ that is highly responsive to inflammatory stimulus. Although further studies are needed to investigate a causal link between neonatal inflammation and programming of pain responsiveness later in life, we can say with a certain degree of confidence that neonatal inflammation is associated with increased pain sensitivity later in life, and this occurs via both peripheral and central immune changes (Figure 1).

Figure 1. Neonatal inflammation can sensitize the pain system for life.

Early life peripheral inflammation affects the development of pain pathways by several mechanisms that work in concert to sensitize the pain system. Immune components are primed in both the periphery and CNS in ways that promote increased pain processing at later stages of development, and increased nerve excitability within the dorsal horn of the spinal cord. Inflammation experienced early in life is thus one of the components that form and fine-tune the developing pain system. Figure is created using images from Servier Medical Art (http://smart.servier.com/).

Another pathway of inflammation-induced long-term pain sensitivity that needs further exploring is via morphological changes in the CNS. Exposure to stress, such as maternal separation, can shape the dendritic and spine morphology in key areas involved in higher cognitive function such as the prefrontal cortex [29]. It is currently not known whether similar changes in dendritic or spine morphology occur following neonatal inflammation. Future studies need to address the impact of neonatal inflammation on neuronal structure, as well as microglial and astrocytic activation, in brain areas involved in the processing of pain during early life-dependent remodeling of neural circuits, as well as the normal development of nerves dependent on immunological signaling molecules. A recent report demonstrated a remarkable astrocyte-microglia communication involved in synapse homeostasis during development. Developing astrocytes release the cytokine IL-33 that drives microglia synapse engulfment in the developing spinal cord and the thalamus. Disruption of IL-33 signaling led to excess synapse and abnormal thalamic circuit formation [19]. Given the modulation of cytokine production centrally following exposure to LPS, it would be interesting to address the impact of this altered cytokine production on the structure of developing nociceptive circuitry.

Peripheral inflammation leads neurophysiological changes in the spinal cord that may affect pain processing

In addition to examination of supraspinal centers associated with altered pain behaviors, spinal populations in the pain pathway have been studied in response to peripheral inflammation. Researchers have demonstrated modulation of inflammatory molecules within spinal cord circuits following peripheral inflammation [30]. However, the question remains as to how these inflammatory molecules impact the physiological output of pain neurons. This is particularly important for the dorsal horn of the spinal cord, where nociceptive and non-nociceptive sensory information interact to modulate the signal passed onto supraspinal centers. The neurophysiology of spinal cord neurons is thus an important contributor to the modulation of nociception and the ultimate experience of pain.

Neurophysiological studies in adult rodents have demonstrated that directly applying proinflammatory molecules, such as IL-6 and TNFα, have significant and varied impacts on spinal neurophysiology [31]. Whilst TNFα increased excitatory synaptic transmission, IL-6 decreased inhibitory synaptic transmission, resulting in an overall increase in excitability of spinal networks. Furthermore, direct application of prostaglandin E2 decreased the input resistance of spinal neurons [32] and enhances long-term potentiation between primary afferent fibres and dorsal horn projection neurons [33], ultimately augmenting the connection between the periphery and the spinal cord. Although these studies provide insight regarding the interactions between inflammatory agents and neurophysiology, studies examining the long-term programming of dorsal horn physiology following neonatal inflammation are limited, and primarily focus on the long-term changes in the adult, at stages well past the resolution of the neonatal insult [34]. Moreover, these existing studies do not examine spinal neurophysiology following inflammation, nor do they examine the trajectory of these changes throughout the critical periods of development. Because of the extensive maturation of both the neural and immune systems during the early postnatal period, it is crucial to examine the bidirectional interactions at critical periods of development.

It has been shown that exposure to LPS at different postnatal ages results in an altered inflammatory soup within the dorsal horn of the spinal cord depending of the timing of infection with regards to critical periods of development. LPS injection at PND 3 caused an upregulation of IL-6, CCL2 and CCL3, whereas these cytokines were down regulated following LPS administration at PND 21 [23]. Recently, a series of studies mapped the neurophysiological properties of superficial dorsal horn (SDH) neurons at three time-points following neonatal exposure to LPS. In summary, these studies demonstrated that neonatal exposure to LPS resulted in subtle changes in intrinsic properties including input resistance [22] and increases in the properties of the rapid-A-type potassium current and network excitability [35•]. Together, this demonstrates subtle, but potentially important, increases in overall excitability within dorsal horn networks. Importantly, these changes were observed as early as three weeks postnatal, at PND 22, much earlier than prior studies focusing ages well into adulthood, which demonstrates that neonatal inflammation has the capacity to alter the developmental trajectory of spinal circuits, as well as the supraspinal centers receiving the output from the dorsal horn of the spinal cord (Figure 1).

The mature pain system is further regulated centrally by peripheral inflammatory activity

Hyperalgesia is one of the core adaptations to sickness, as discussed in the second section of this manuscript, along with decreased locomotor activity, lethargy, anorexia, depressive and anxiety-like behavior, and fever [10,21]. The spinal neuroimmune mechanism in inflammation-driven experimental hyperalgesia in adult rodents have been described in detail before [10,11,25,36], and we will, therefore, limit this section in this review. Some of the core features include microglia activation via toll-like receptor (TLR) activation and immune cell infiltration to the CNS [36,37]. This line of research shows that the priming of microglia by peripheral inflammatory signals may be a key mechanism in the transition between acute to chronic pain.

Experimental inflammation and pain sensitivity in humans

Rodent LPS experiments allow translational conclusions, but these should be made with caution. When human participants are injected with LPS, significantly lower doses are used that induce milder sickness responses. Rodent LPS doses can be up to 100 times higher, partly due to a higher tolerance for endotoxins in rodents, but also to induce clearly detectable behavioral changes. However, pain sensitivity appears readily affected by immune provocation in humans as well, rendering healthy humans more pain sensitive during the acute phase of the systemic inflammation. Hyperalgesia is accompanied by an increase in subjective anxiety [38], depressed mood [39] and increased fatigue [40] and motivational changes [6], much like in animal models. Deep pain, such as pressure pain and visceral pain, is affected at lower doses of LPS than cutaneous and mechanical pain (superficial pain) [41,42]. Enhanced somatic pain and visceral pain sensitivity usually coincide with higher circulating levels of pro-inflammatory cytokines. Cerebrospinal fluid (CSF) cytokine levels or inflammation-related markers in the brain have not been studied during experimental pain testing specifically, but CSF IL-6 levels seem to correlate with inflammation-induced negative mood [43]. Two studies have shown changes in pain-induced brain activity, as induced by visceral, cutaneous or pressure pain, during acute inflammation, in areas involved in the affective component of pain [44], such as the amygdala, cingulate and prefrontal cortices [42,45]. Furthermore, descending pain inhibition is attenuated by inflammatory activity [41,45]. This pattern is similar to that discussed earlier in rodent neonatal LPS studies. Peripheral inflammation appears to, in both situations, increase pain sensitivity through changes in brain areas involved in affective and regulatory functions of pain perception. Several neuroimmune pathways working together to affect pain perception during peripheral inflammation are thus plausible; via peripheral nerve sensitization as traditionally argued [37], via modulations in the spinal cord as discussed previously, and via functional changes in the pain network in the brain (Figure 2).

Figure 2. Peripheral inflammation modulates the pain system throughout life.

The figure depicts a model that incorporates findings from rodent and human experimental inflammation models in different developmental stages. In this model, peripheral inflammation modulates pain processing from birth to adulthood. Early life inflammation impacts the development of an individual’s pain system via immunological and neuronal changes, that in turn define individual differences in neuroimmune communication later in life. A primed pain system is more susceptible to the detrimental effect of immune-to-brain signaling during illness or low-grade inflammation. Augmented periphery-to-CNS neuroimmune interaction may result in the development, maintenance and spreading of long-term pain, as well as in the development of comorbid syndromes that are commonly experienced by chronic pain patients. Figure is created using images from Servier Medical Art (http://smart.servier.com/).

Inflammation in clinical pain populations

Models of acute inflammation are useful to study the specific effect of inflammation on the nervous system, without bias introduced by comorbid disease, inflammatory progress over time or different types of inflammation as present in clinical populations. Interestingly, peipheral inflammation is related to increased pain sensitivity in the general population as well, regardless of confounders or even concurrent chronic pain [46]. Experimental inflammatory stimuli must be mild and transient, and only healthy adults are studied. Chronic inflammation on the other hand, has a different composition of inflammatory components than the acute innate inflammatory response induced by LPS, and may change over time. Studies on central neuroimmune interactions driven by chronic peripheral inflammation in humans is limited, but a recent meta-analysis of both experimental and observational inflammatory studies suggests global effects on brain function associated with peripheral inflammation [7••]. Inflammation consistently affected areas of the limbic system, basal ganglia, brain stem and prefrontal cortex, which are areas implicated in pain processing [44]. In a chronic pain cohort, the inferior perietal lobule was recently identified as an inflammation-sensitive hub in the brain [47]. Recent large-scale peptide analyses suggest ongoing inflammatory activity both centrally [48,49] and peripherally [49,50] in different chronic pain populations. Reported cytokine levels and types vary between study populations and are sometimes inconsistent [51•]. A recent study on low-back pain suggests difference in the neuroimmune crosstalk from the periphery to the CNS depending on the etiology of the back pain. While peripheral IL-8 levels correlated with central levels (as measured in CSF) for patients with lumbar disc herniation [51•], such an association was instead found for CCL2 in degenerative disc disease. Proteomics studies of this kind are needed to identify the specific networks of inflammatory peptides related to a specific pain state, and how the peripheral inflammatory soup relates to central inflammatory activity. However, not only ongoing low-grade inflammation may be of importance, but the immunoreactivity of pain patients as well, in a similar manner as discussed in the neonatal rodent models. In chronic pelvic pain patients, IL-6 and IL-1β levels expressed by isolated peripheral blood mononuclear cells after in vitro LPS stimulations were higher in pain patients compared to healthy controls [52]. The inflammatory reactivity was associated with reported pain levels and, interestingly, with comorbid as well as widespread pain. These associations were not seen for plasma IL-6, but plasma levels did correlate with experimental pressure pain thresholds, which corresponds to findings from in vivo human LPS models. Most importantly, LPS stimulated inflammatory activity was associated with less improvement of symptoms over time [53••]. These findings suggest that individuals with chronic pain may have a primed immune system [54] that plays a role in their disease development over time. This priming of immune components occurs early in life and continues during later developmental stages, rendering the pain system vulnerable to future inflammatory insults and pain chronification, maintenance and proliferation (Figure 2).

Conclusion

Peripheral inflammation sensitizes the central pain system. An early life inflammation can cause permanent changes in immune reactivity peripherally and centrally, paralleled with neurophysiological changes in the spinal cord. Inflammation thus shapes the pain system early in life, and can program pain sensitivity throughout development and adulthood. In the adult organism, similar neuroimmune interactions may be repeated. Peripheral inflammation is propagated into central inflammation and alters the neural activity of brain areas that encode affective and regulatory components of pain perception. The activated immune system thus affects pain sensitivity, mood and motivation in humans. While some mechanisms for functional changes of the pain system after neonatal inflammation have been identified, structural changes are also plausible. The nervous system also appears to be sensitive to the timing of inflammation, with outcomes of the inflammatory insult depending on the developmental window. Studying how the developing pain system interacts with inflammatory cues is crucial for understanding individual differences in chronic pain development and treatment success. Several chronic pain disorders with different etiologies have been associated with peripheral and central low-grade inflammation. Recent research is attempting to identify inflammatory signatures specific for clinical subpopulations, or overarching for disparate chronic pain populations, and understand their biological mechanisms in the disease, and prognostic and predictive utility. Understanding how peripheral inflammation may drive and maintain pain and comorbid disorders gives opportunities for new therapeutic approaches.