INVITED REVIEWS
Senso-immunology: The Emerging Connection between Pain and Immunity
2023 Volume 72 Issue 3 Pages 77-87
Details
2023 Volume 72 Issue 3 Pages 77-87
The sensory and immune systems have been studied independently for a long time, whereas the interaction between the two has received little attention. We have carried out research to understand the interaction between the sensory and immune systems and have found that inflammation and bone destruction caused by fungal infection are suppressed by nociceptors. Furthermore, we have elucidated the molecular mechanism whereby fungal receptors are expressed on nociceptors and skin epithelium, how they cooperate to generate fungal pain, and how colitis and bone metabolism are regulated by mechanosensors expressed on the gut epithelium. Recently, we found that nociceptors prevent septic death by inhibiting microglia via nociceptor-derived hormones. This review summarizes our current state of knowledge on pain biology and outlines the mechanisms whereby pain and immunity interact. Our findings indicate that the sensory and immune systems share a variety of molecules and interact with each other to regulate our pathological and homeostatic conditions. This prompted us to advocate the interdisciplinary science named “senso-immunology,” and this emerging field is expected to generate new ideas in both physiology and immunology, leading to the development of novel drugs to treat pain and inflammation.
In the medical community of the twentieth century, pain associated with trauma and infection was considered a diagnostic clue and therefore something that did not require treatment. Hence, investigations on the molecular biology of acute pain stopped at an immature stage. However, in 1997, the discovery of the transient receptor potential vanilloid 1 (TRPV1) channel, which is expressed on nociceptors and senses pain, led to the cloning of many ion channels involved in pain perception and a dramatic advance in pain biology.1 In 2010, Piezo family genes were identified as ion channels that sense mechanical stimuli,2and researchers in various fields are currently working on the functional analysis of these genes. In this context, research began on infection-associated nociception, which had previously remained unexplored.3 Consequently, the direct recognition of pathogens by nociceptors, as if they were innate immune cells, was discovered. Notably, nociceptors have not only been reported to act as exacerbators and protective factors in various chronic diseases via the secretion of neuropeptides but have also been suggested to regulate bacterial or fungal infection.4 In this review, we discuss the current state of knowledge on pain biology and outline the hidden mechanisms whereby pain is associated with immune phenomena and their biological functions, based on the results of our recent research.
TRP channels are activated by various pain-related compounds and are primarily expressed in sensory neurons. There are six subgroups of TRP channels involved in nociception: TRPV, TRP canonical (TRPC), TRP mucolipin (TRPML), TRP polycystin (TRPP), TRP ankyrin (TRPA), and TRP melastatin (TRPM). It is well known that sensory neurons contain three types of fibers: Aβ-, Aδ-, and C-fibers. Aβ-fibers are large-diameter myelinated fibers that respond to touch and pressure. Aδ-fibers are small-diameter myelinated fibers that respond to sharp pricking pain. C-fibers are small, unmyelinated fibers responsible for slow pain, such as aching and burning pain. TRP channels have a structure that consists of six transmembrane helical protein domains with intracellular C- and N-termini. The channel pore formed by the six transmembrane structures regulates the movement of cations. TRP channel activity is regulated by G protein-coupled receptor signaling and phospholipase C (PLC) activation. These biochemical processes are strongly influenced by changes in the cellular environment. Specifically, most TRP channels respond to extracellular and intracellular stimuli such as changes in temperature, osmolarity, pH, and concentrations of various chemicals.1 TRPV1, which is expressed in C-fibers of primary afferents and is activated by protons, heat, and capsaicin, a component of chili peppers, is particularly important in detecting pain associated with trauma and burns.5 One TRP channel that has been studied as much as TRPV1 is TRPA1, which is activated by cold stimuli, mustard oil, and allyl isothiocyanate, a component of horseradish. Most TRPA1 expressed on C-fibers colocalizes with TRPV1. Notably, TRPA1 expression increases during inflammation, and it has been suggested that TRPA1 may be closely involved in the development of inflammatory pain.1,6
Bacterial and fungal infections of tissues where nociceptors are located induce unpleasant sensations, such as pain and itching. Activation of the nociceptive system by pathogen infection has been vaguely assumed to be caused by the immune system-mediated inflammatory response and TRP channels, and the molecular mechanism of this phenomenon has not been thoroughly investigated. Recently, it was reported that pain associated with Staphylococcus aureus infection is induced by α-hemolysin, a hemolytic toxin secreted by S. aureus.7 Bacterial infection induces the production of pro-inflammatory cytokines by activating Toll-like receptors (TLRs),8 but the pain associated with S. aureus infection also occurs in MyD88-deficient mice, in which TLR signaling is almost abolished. In contrast, mice infected with S. aureus lacking α-hemolysin show little pain when compared with mice infected with wild-type bacteria. In addition, in vitro stimulation of nociceptors with α-hemolysin elicits calcium signaling and action potentials.7 These results suggest that pro-inflammatory cytokines are not involved in the development of pain associated with bacterial infection, and that their role lies in the direct stimulation of nociceptors by bacterial hemolytic toxins. To date, there has been almost no research on the physiological function of infection-associated pain. It remains unclear whether pain can modify the degree of inflammation and tissue damage that accompanies infection in the host. To elucidate the physiological function of pain, we began a decade ago to investigate a new interdisciplinary field that combined physiology and immunology.
To elucidate the unknown functions of nociceptors, it is essential to create an animal model that is deficient in nociceptors. By crossing Nav1.8 Cre mice with Rosa26 DTA mice, we generated nociceptor-null mice in which C-fibers expressing Nav1.8 ion channels were congenitally defective throughout the body.9 The mice did not jump when placed on a steel plate at 65 °C for more than 1 min and did not show any signs of resistance to mechanical stimulation of the plantar surface of their hind legs with thin filaments, suggesting that they lost sensation to heat and mechanical stimulation. Subcutaneous injection of lipopolysaccharide (LPS), a component of gram-negative bacteria,8 into the plantar feet of nociceptor-null mice induced a rapid inflammatory response, resulting in swelling of the entire foot, but the extent of the swelling was not different from that in wild-type mice. The interleukin 6 level in the blood after subcutaneous injection of LPS was also similar to that in wild-type mice, suggesting that nociceptors are not involved in the inflammatory response to LPS.9 When Candida albicans, a typical commensal fungus, was injected subcutaneously into the plantar feet of wild-type mice and nociceptor-null mice, only slight foot swelling was observed in wild-type mice, whereas severe swelling of the entire foot and lysis of the calcaneus bone were observed in nociceptor-null mice. Furthermore, when β-glucan, a well-known fungal cell wall component of C. albicans,8 was administered to the plantar feet of mice, wild-type mice showed only slight swelling of the feet, whereas nociceptor-null mice showed swelling of the entire foot and lysis of the calcaneus bone.9 These results indicate that the Nav1.8-positive nociceptor does not inhibit gram-negative bacterial inflammation but inhibits fungal osteo-inflammation (Fig. 1).
Nociceptor inhibits fungal osteo-inflammation.
Candida albicans-derived soluble β-glucan (CSBG) stimulates nociceptors via the Dectin1–TRPV1/TRPA1 axis leading to the production of calcitonin gene-related peptide (CGRP). CGRP inhibits pro-inflammatory cytokine production and osteoclast fusion.
Candida albicans is a commensal fungus that lives on the skin and inside the vagina and is not normally harmful to the human body.10 However, when the homeostasis of the immune system and flora is disturbed by the administration of anticancer drugs or antibiotics, C. albicans begins to proliferate, causing a painful and itchy rash on the oral mucosa, external auditory canal, pubic region, and interdigital areas.11 In addition, although less frequent, painful osteomyelitis has been reported after orthopedic surgery in immunocompromised patients.12 However, the mechanism whereby C. albicans causes these uncomfortable symptoms of opportunistic infections is unknown. Recently, C. albicans was shown to secrete a pore-forming protein called candidalysin, which can damage epithelial cells at the site of infection.13 When C. albicans lacking candidalysin and wild-type C. albicans were administered to mice for the analysis of pain behavior, there was no difference in their responses. Furthermore, no obvious increase in intracellular calcium concentration was observed after the stimulation of nociceptors extracted in vitro with candidalysin.14 These results indicated that candidalysin, a pivotal virulence factor of C. albicans, is not the cause of pain induced by C. albicans. Generally speaking, C. albicans becomes dormant in a harsh environment by developing a rounded shape called the yeast form, but when exposed to a humid environment at 37 °C, it converts to a string-like form called the mycelium form and becomes actively proliferating.15 We analyzed the culture supernatant during this transformation process, assuming that a pain-inducing substance was present among the substances secreted by C. albicans. Notably, we found that soluble β-glucan (C. albicans-derived soluble β-glucan, CSBG) was released from the fungus in large amounts during transformation.14 It is well known that CSBG induces pro-inflammatory cytokine production by activating Dectin-1 receptors expressed on macrophages.16 Surprisingly, analysis using calcium imaging revealed that functional Dectin-1 is expressed not only in macrophages, but also in nociceptors and skin epithelial cells. Subcutaneous administration of CSBG in mice induces transient pain that disappears within a short period of time, but this was not observed in mice lacking Dectin-1, suggesting that the acute transient pain response induced by CSBG may be caused by Dectin-1 stimulation of nociceptors.14 Furthermore, CSBG-treated mice showed allodynia, a reduction in pain threshold (feeling pain from stimuli that are not normally painful),17 for 24 h after the acute transient pain response disappeared. Interestingly, this phenomenon was also observed in mice lacking Bcl-6 or Malt-1 genes, which are unable to produce pro-inflammatory cytokines in response to CSBG.14,18 These results clearly suggest that allodynia induced by CSBG is not caused by pro-inflammatory cytokines. Recently, adenosine triphosphate (ATP) has attracted attention as the causative agent of neuropathic pain.19 We evaluated CSBG-induced allodynia levels in mice after the administration of A317491, an ATP receptor (P2X receptor) inhibitor, and observed significantly lower allodynia levels than in the control group.14 This result suggests that fungal infection-associated pain may be triggered by ATP. Recently, it was reported that ATP is stocked in the cytoplasm in the form of secretory granules by the action of the vesicular nucleotide transporter (VNUT).20 Surprisingly, CSBG-induced allodynia was completely abolished in mice lacking the VNUT gene, which is unable to produce secretory granules of ATP, and in mice double deficient in the TRPV1 and TRPA1 genes.14 This means that CSBG induces allodynia by activating Dectin-1 signaling in skin epithelial cells, which releases ATP, which in turn activates ATP receptors (P2X receptors) expressed on nociceptors, and that TRPV1 and TRPA1 are essential for the induction of allodynia by ATP. In addition, mice treated with clodronate, an osteoporosis drug that inhibits VNUT activation,21 showed a marked improvement in CSBG-induced allodynia and itching, indicating that VNUT is a therapeutic target for eliminating fungal infection-associated symptoms (Fig. 2).14
Mechanism of fungal pain generation.
CSBG directly stimulates nociceptors via Dectin1 to generate acute pain. CSBG also induces allodynia. CSBG-induced allodynia is not dependent on the immune system, but instead on keratinocyte-derived ATP.
The exacerbation of fungal osteo-inflammation in nociceptor-null mice was alleviated by tumor necrosis factor alpha (TNF-α) neutralizing antibodies, indicating that Nav1.8 positive nociceptors prevent osteo-inflammation by suppressing TNF-α production during fungal infection.9 To identify the humoral factors that nociceptors release during fungal infection to suppress osteo-inflammation, we collected serum from C. albicans-infected mice over time and measured the levels of neuropeptides known to be expressed in nociceptors. The results showed that the concentration of the neuropeptide calcitonin gene-related peptide (CGRP)22,23 was specifically elevated after infection with C. albicans.9 CGRP concentration in blood showed a transient increase 24 h after subcutaneous injection of LPS but was not detected thereafter. However, after subcutaneous injection of CSBG, the CGRP concentration in blood remained very high for several days. The increase in blood CGRP concentration after subcutaneous injection of CSBG was abolished in Dectin-1-deficient, TRPV1 and TRPA1 double-deficient, and nociceptor-null mice. In contrast, mice lacking B-cell lymphoma/leukemia 10 (Bcl10) and mucosa-associated lymphoid tissue lymphoma translocation protein 1 (Malt1), which are essential for Dectin-1 signaling in immune cells,24 showed normal CGRP induction.9 These results suggest that subcutaneous injection of CSBG stimulates Dectin-1 in nociceptors and induces CGRP production by activating TRPV1/TRPA1 in nociceptors through an unknown signaling mechanism. Previous studies using myeloid lineages have shown that signaling downstream of Dectin-1 is mediated by PLC activation.25 Therefore, we examined the phosphorylation of PLC family proteins in the CSBG-stimulated nociceptors. Notably, western blot analysis of nociceptors stimulated with CSBG showed enhanced phosphorylation of PLC-γ2. Furthermore, administration of U73122, a PLC inhibitor, prevented the increase in blood CGRP concentration induced by the subcutaneous injection of CSBG.9 These results indicate that CGRP is induced in the nociceptor by a unique pathway that is not mediated by Bcl10 or Malt1 through the activation of TRPV1/TPRA1 via the Dectin-1-PLC axis. Interestingly, TRPV1 and TRPA1 double-deficient mice showed a prominent exacerbation of CSBG-induced osteo-inflammation. Given that administration of CGRP also caused remission of CSBG-induced osteo-inflammation in nociceptor null mice and TRPV1 and TRPA1 double-deficient mice, we speculated that the exacerbation of fungal osteo-inflammation in these mice was caused by the reduced production of CGRP.9 Further analysis revealed that osteoclasts simultaneously stimulated with CGRP and the osteoclast differentiation inducer receptor activator of nuclear factor kappa-Β ligand (RANKL)26,27 showed a rapid increase in intracellular cAMP levels, which promoted actin depolymerization. Interestingly, the osteoclasts induced by CGRP and RANKL showed normal expression of differentiation markers but were mononuclear because of impaired cell fusion. Because multinucleation by cell fusion is essential for osteoclast activity,28 CGRP is considered a bone-protective peptide that targets cell fusion. Notably, culture supernatants of nociceptors markedly inhibited RANKL-induced multinucleation of osteoclasts, but when the inhibitor of CGRP, Olcegepant, was added to the culture supernatant, this phenomenon was no longer observed.9 These results indicate that nociceptor-derived CGRP, through conjugation with RANKL signaling, increases intracellular cAMP levels and prevents osteoclast fusion. Interestingly, CGRP inhibited TNF-α production from macrophages stimulated with mannan, a TLR ligand that constitutes the C. albicans cell wall, by only 10%–20%, whereas it inhibited TNF-α production from CSBG-stimulated macrophages by more than 70%.9 These facts indicate that CGRP is a neuropeptide that suppresses Dectin-1 signaling more potently than TLR signaling. We screened for transcription factors expressed in myeloid cell lineages that were not induced by stimulation with mannan + CGRP but were induced by stimulation with Dectin-1 ligand + CGRP. We found that the transcription factor Jun dimerization protein 2 (Jdp2)29,30,31 was strongly induced by stimulation with Dectin-1 ligand + CGRP. Detailed analysis revealed that the transcription factor Jdp2 induced by stimulation with Dectin-1 ligand + CGRP suppressed NF-κB activity by directly binding to the p65 subunit of NF-κB, leading to the anti-inflammatory effect of CGRP. Suppression of CGRP-stimulated CSBG-induced inflammation was completely abolished in macrophages derived from Jdp2-deficient mice.9 These results indicate that Nav1.8-positive nociceptors produce large amounts of CGRP via Dectin-1 during fungal infection and that CGRP suppresses osteoclast fusion and macrophage inflammation via induction of cAMP and the transcription factor Jdp2 (Fig. 1).
Witnessing a blindfolded street performer walk a tightrope sometimes causes a marked drop in blood pressure because of the vagus nerve reflex. However, trained street performers rarely fall from their ropes while blindfolded, and surprised spectators do not die of shock. To walk a tightrope blindfolded and not die from the vagus nerve reflex, we must assume the presence of sensors that detect the force exerted by the rope on the soles of our feet and the blood pressure exerted on our blood vessels. For organs and tissues to sense and respond to pressure, they must have receptors that convert mechanical stimuli to biological signals; however, such receptors have remained unknown until recently. In 2010, Patapoutian and colleagues demonstrated in mice and Drosophila that cation channels of the Piezo family are responsible for this mechanism by permeating calcium in response to changes in membrane tension, revealing, for the first time, the full molecular picture of mechanosensing.2,32 The Piezo family includes Piezo1 and Piezo2; the former of which has been reported to be involved in the proper orientation of vascular and lymphatic endothelium and blood pressure sensing,33,34,35,36,37 whereas Piezo2 is implicated in the sensing of pressure and mechanical stimuli in nociceptors.38,39 Recently, it was reported that humans with a gain-of-function mutation in Piezo1 are prone to hemolysis because of erythrocyte volume reduction mediated by excessive water and potassium efflux through the opening of calcium-activated potassium channels.40,41 Furthermore, Piezo1 has also been shown to be essential for mechanostimulation-evoked chemokine expression in macrophages, and it is now understood that Piezo channels play an important role in immune cell migration.42
The discovery of the Piezo family has greatly advanced our understanding of pain in response to mechanical stimuli; however, the role of this ion channel in physiological functions such as inflammation, peristalsis, and bone metabolism remains unclear. We found that Piezo1 is expressed in the intestinal epithelium and bone-metabolizing cells, such as osteoclasts and osteoblasts.43 To elucidate the role of Piezo1 in the intestine and bone, we developed mouse models of intestinal epithelium-specific Piezo1 deficiency (Villin Cre-Piezo1 flox/flox), osteoclast-specific Piezo1 deficiency (LysM Cre-Piezo1 flox/flox), and osteoblast-specific Piezo1 deficiency (Col1a1 Cre-Piezo1 flox/flox) and analyzed their phenotypes. The results showed that the bone mass of osteoclast-specific and osteoblast-specific Piezo1-deficient mice was normal, whereas that of intestinal epithelium-specific Piezo1-deficient mice was markedly increased. Bone morphometry revealed that osteogenesis by osteoblasts was enhanced in the bones of intestinal epithelium-specific Piezo1-deficient mice, which was thought to be the cause of increased bone mass. In vivo intestinal peristalsis was observed with the Shimadzu SAI-1000 imaging system (Shimadzu, Kyoto, Japan) after oral administration of a near-infrared fluorescent probe, and intestinal peristalsis in mice lacking intestinal epithelium-specific Piezo1 was markedly reduced when compared with wild-type mice. In addition, when intestinal inflammation was induced in intestinal epithelium-specific Piezo1-deficient mice by administering dextran sulfate sodium (DSS) at a concentration that killed wild-type mice in 1 week, all mice survived. Histological analysis of the colon revealed that mice lacking intestinal epithelium-specific Piezo1 did not show DSS-induced colitis. These results indicated that Piezo1 in the intestinal epithelium negatively regulates bone formation, promotes intestinal peristalsis, and exacerbates intestinal inflammation.43 Using transcriptome analysis, we found that the levels of serotonin, a hormone known to promote intestinal peristalsis,44 exacerbate intestinal inflammation,45 and inhibit bone formation,46 were decreased in the intestinal epithelium lacking Piezo1. Therefore, when intestinal epithelium-specific Piezo1-deficient mice were treated with serotonin for 1 month, intestinal peristalsis, DSS-induced colitis, and bone formation rates were comparable with those in wild-type mice. To test whether Piezo1 induces serotonin production in response to mechanical stimulation by intestinal peristalsis, we applied in vitro stretch contraction stimulation to the intestinal epithelium using the STREX system (Strex, Osaka, Japan). Contrary to our expectations, the wild-type and Piezo1-deficient intestinal epithelia showed similar levels of serotonin production. Therefore, we speculated that Piezo1 in the intestinal epithelium is not activated in response to mechanical stimulation. To determine whether intestinal Piezo1 produces serotonin in an intestinal bacteria-dependent manner, we reduced intestinal bacteria by administering a cocktail of antibiotics to mice and found that intestinal epithelial serotonin production was suppressed and bone mass was increased in wild-type mice, whereas this phenomenon was not observed in intestinal epithelium-specific Piezo1-deficient mice. These results suggest that some intestinal bacteria-derived molecules in the feces may serve as ligands for Piezo1.43 In fact, fecal lysate activated Piezo1, suggesting that this hypothesis is highly probable. We then performed calcium imaging of three fractions of fecal lysate: protein, DNA, and RNA fractions, and observed a calcium response only in the RNA fraction, indicating that fecal RNA can activate Piezo1. Furthermore, the ligand activity of the fecal RNA fraction was abolished by RNase A, which degrades single-stranded RNA (ssRNA), indicating that ssRNA is the ligand for Piezo1.43 ssRNAs are known to induce pro-inflammatory cytokine production in macrophages through the TLR7-MyD88 pathway.8 However, when ssRNA was sprinkled on intestinal epithelia lacking TLR7 or MyD88, serotonin production was comparable with that of the wild-type. Given that calpain and Akt signaling are known to be activated when intracellular calcium concentration increases,33,47 treatment of intestinal epithelium with the calpain inhibitor PD150606 or the Akt inhibitor AZD5363 markedly inhibited serotonin production in response to ssRNA.43 These results indicate that ssRNA induces serotonin production via the calpain and Akt pathways rather than the TLR pathway. Finally, to clarify the physiological significance of ssRNA in the intestinal tract, we performed a colonic infusion experiment with RNase A. The results showed that colonic RNase A-infused mice had increased bone mass and decreased intestinal peristalsis, accompanied by a decrease in blood serotonin concentration. These results suggest that fecal ssRNA acts as a serotonin-inducing molecule via Piezo1 (Fig. 3).43
Mechanism of Piezo1-mediated gut and bone homeostasis.
Fecal ssRNA activates gut Piezo1, leading to the generation of serotonin. Gut-derived serotonin regulates the level of bone formation, peristalsis, and gut inflammation.
Our studies have revealed that nociceptors suppress fungal osteo-inflammation via the CGRP-Jdp2 axis9,14 and that the mechanosensor Piezo1 expressed in the gut regulates bone mass and colitis by recognizing ssRNA from intestinal bacteria.43 This prompted us to investigate whether the pathogenesis of bacterial sepsis, which is estimated to account for 20% of all deaths worldwide,48 is modulated by nociceptors. Bacterial sepsis is an infectious disease with a systemic inflammatory response, and 30% of patients die without treatment. Various studies have suggested that the cause of death is multi-organ failure with abnormal temperature, tachycardia, hypotension, and coagulopathy caused by cytokine storms. However, the administration of neutralizing antibodies against pro-inflammatory cytokines does not improve the mortality rate.49 This suggests that the pathophysiology of sepsis is unclear. We have been conducting osteo-immunological analysis of nociceptor null mice in which Nav1.8-positive nociceptors were selectively removed, and found that these mice were vulnerable to LPS-induced shock despite the absence of abnormal pro-inflammatory cytokine production, and died suddenly with convulsions about 36 h after LPS injection.50 Fluorodeoxyglucose (FDG)-positron emission tomography (FDG-PET) imaging and brain metabolome analysis revealed impaired brain FDG accumulation and decreased brain ATP concentration after systemic LPS injection as well as a significant decrease in the phosphorylation of brain hexokinase 1, which plays a key role in the initiation of cellular respiration. Given that the occurrence of seizures suggests damage to the central nervous system, we hypothesized that the direct cause of septic death is in the brain and that nociceptors prevent septic death by producing molecules responsible for brain protection. In fact, it has been reported that administration of the antiepileptic drug valproic acid in septic mice improves mortality.51 Analysis of brain metabolome pathways in LPS-injected wild-type and nociceptor-null mice suggested that dramatic activation of the kynurenine pathway occurs just prior to death and that nociceptors may suppress this phenomenon. Specifically, the expression of indoleamine-2,3-dioxygenase 1 (IDO1), an enzyme involved in the quinolinic acid synthesis pathway expressed in microglial cells,52 is upregulated just before septic death, resulting in a rapid increase in the brain concentration of quinolinic acid, which simultaneously overexcites neurons and impairs the phosphorylation of hexokinase 1 in neurons. Intrathecal injection of quinolinic acid immediately killed mice, accompanied by impaired phosphorylation of brain hexokinase 1, while intrathecal injection of the IDO1 inhibitor 1-methyl-D-tryptophan rescued wild-type and nociceptor-null mice from septic death.50 Based on these data, we hypothesized that during sepsis, nociceptors release humoral factors into the blood that inhibit IDO1 expression and that this factor enters the brain and acts on microglial cells to prevent septic death. Transcriptome analysis of dorsal root ganglion cells from wild-type mice and nociceptor-null mice after systemic LPS injection revealed that the expression of the C-type lectin family protein regenerating islet-derived protein 3 gamma (Reg3γ), a type of antimicrobial peptide,53 was lost in the latter, while the expression of exostosin-like glycosyltransferase 3 (Extl3), a receptor for this protein,54 was more prominent in the brain microglial cells. When mice specifically lacking Reg3γ in nociceptors were created, the expression of Reg3γ in the blood was lost; at the same time, the mice became extremely vulnerable to sepsis, and the metabolic state of the brain after the onset of sepsis was a phenocopy of that of nociceptor-null mice.50 Stimulation of microglial cells with Reg3γ resulted in the cessation of quinolinic acid production by suppressing LPS-induced IDO1 expression. In addition, intrathecal injection of Reg3γ into wild-type mice prevents septic death. In conclusion, nociceptors release Reg3γ as a brain-targeted hormone during sepsis, which inhibits IDO1 expression in brain microglial cells, thereby preventing septic death (Fig. 4).50
Mechanism of nociceptor-mediated tolerance.
Lipopolysaccharide (LPS)-stimulated nociceptor produces Reg3γ. Reg3γ suppresses the expression of microglial indoleamine-2,3-dioxygenase 1 (IDO1), leading to protection from LPS-induced death.
Our research revealed that nociceptors modify the pathogenesis of fungal osteo-inflammation9 and bacterial sepsis in mice50; however, whether human nociceptors have a similar function remains unclear. To clarify this point, it is necessary to unravel the nature of Congenital Insensitivity of Pain with Anhidrosis (CIPA), a genetic disorder in which nociceptors are defective.55 CIPA is caused by a loss-of-function mutation in the nerve growth factor receptor gene neurotrophic receptor tyrosine kinase 1 (NTRK1), which was first identified by Dr. Yasuhiro Indo of Kumamoto University.56 Because CIPA causes both nociceptor and sympathetic nerve defects, it is not possible to conclude that the phenotype is the result of a nociceptor defect. However, CIPA is frequently associated with osteomyelitis, fractures, and trauma; it is possible that many of these traits may be explained by a deficit in nociceptors.57 Whether our findings in nociceptor-null mice are valid in humans will soon be clarified through studies on the relationship between CIPA and fungal infections from an orthopedic perspective. In addition, children with CIPA sometimes die of sepsis triggered by trauma, but the degree of inflammatory reaction is not strong (personal communication from Dr. Yasuhiro Indo). As mentioned above, when sepsis is induced in nociceptor-null mice, all mice die within approximately 36 h, even though their systemic inflammatory cytokine levels are unchanged compared with those in wild-type controls.50 These findings suggest that the vulnerability to sepsis of patients with CIPA is caused by an unknown mechanism that cannot be explained by inflammatory mechanisms, and that if the cause of septic death in these patients is a defect in nociceptor-derived Reg3γ, administration of this protein may prevent septic death.50 Our research group is currently working on this aspect of the study.
Although the sensation associated with pathogen infection is a familiar phenomenon, it has been a neglected topic in the field of medical research until now. Sensory receptors are expressed not only in neurons but also in the intestinal epithelium, as in the case of Piezo1, but research to clarify the immune functions of these receptors expressed in tissues other than sensory organs has only just begun. We believe that reexamining the immune system from the perspective of sensory biology may contribute to the creation of a new interdisciplinary field called “senso-immunology” (Fig. 5).4,58 It is important for physiologists and immunologists to cooperate with each other to promote senso-immunology to develop medical treatments that can improve the complications and prognosis of various inflammatory or painful conditions.
Senso-immunology: the novel research field.
Understanding the interplay between sensory and immune systems may open new avenues for the development of an interdisciplinary field called “senso-immunology.”
The author thanks T. Kondo for fruitful discussions. The author also acknowledges assistance from the following organizations for funding the research discussed in this article: Takeda Science Foundation, Japan Society for the Promotion of Science (KAKENHI JP21H03114, JP16K15665, JP18H02970, JP16K15665, JP19K22712), Translational Research Network Program of the Japan Agency for Medical Research and Development, Takeda Science Foundation, Lotte Foundation, Kowa Life Science Foundation, Kanazawa Medical Research Foundation, Akashi Medical Foundation, Yakult Bioscience Research Foundation, Ichiro Kanehara Foundation for the Promotion of Medical Sciences and Medical Care, Japan Dairy Association, Lydia O’Leary Memorial Pias Dermatological Foundation, Canon Foundation, Japan Intractable Diseases Research Foundation, Uehara Memorial Foundation, Terumo Life Science Foundation, Inoue Foundation for Science, Brain Science Foundation, Nakajima Foundation, Life Science Foundation, Mitsubishi Foundation, Astellas Foundation, Mochida Memorial Foundation, Gushinkai Foundation, an Okamoto Research Award, and Toyoaki Scholarship Foundation.
The author declares no conflict of interest.
