Graphical abstract graphic file with name fx1.jpg [33]Open in a new tab Highlights * • Stimulation of TRPV1+ peripheral nerve attenuates systemic inflammation * • Somato-autonomic reflex induces the secretion of catecholamines * • Activating TRPV1+ afferent at nape drives sympathetic and vagal efferent pathways * • TRPV1+ somatosensory afferent stimulating modulates the expression of splenic genes __________________________________________________________________ Neuroscience; Sensory neuroscience Introduction Inflammation can eliminate invading pathogens and restore physiological homeostasis, but an excessive or chronic inflammatory response damages various tissues and induces multiple diseases. To date, effective prevention or treatment methods are lacking. Moxibustion and apitherapy, which can produce anti-inflammatory responses and analgesia and treat a wide range of diseases, have long been used in traditional oriental medicine, but their therapeutic mechanism remains unclear. Moxibustion-like stimulation (>43°C) can activate peptidergic C-fibers in the skin.[34]^1 Ginger-partitioned moxibustion or garlic-partitioned moxibustion, in which gingerol and allicin are agonists of TRPV1, can increase the therapeutic effects. Melittin, the main peptide in bee venom, activates TRPV1 ion channels and TRPV1+ peripheral sensory nerves.[35]^2 Therefore, we speculate that stimulation of TRPV1+ peripheral afferents may modulate the inflammatory response. TRPV1, a nonselective cation channel, is a thermoreceptor activated by noxious heat (>43°C) and some endogenous and exogenous mediators.[36]^3^,[37]^4 TRPV1 is highly expressed in dorsal root ganglia (DRG) and nodose ganglion (NG), which transmit somatosensory and vagal sensory information, respectively. Based on single-cell RNA sequencing (RNA-seq) and function assay, the sensory neurons of DRG are divided into 11 distinct cell types. TRPV1 is selectively expressed in NP2, NP3, and PEP1 subgroups of nociceptors.[38]^5^,[39]^6^,[40]^7 Our previous studies indicated that the percentage of TRPV1+ nociceptors in total DRG neurons is approximately 20%, and TRPV1+ primary afferents project to lamina I and outer layer of lamina II within the dorsal horn of the spinal cord.[41]^8 Noxious and/or thermal stimuli from the skin and deep tissues are further transmitted to several brain regions. NG contains 18 types of vagal neurons. TRPV1 is specifically expressed in NG12–NG16 neuron subtypes that transmit harmful or inflammatory signaling of internal organs to the nucleus of the solitary tract (NTS) in the brainstem.[42]^9 Then, the autonomic efferent nerves are activated to modulate visceral functions. Accumulating evidence indicates that electric stimulation of nerves can regulate peripheral immune responses. Electric stimulation of vagal afferent and efferent nerves can attenuate systemic inflammation.[43]^10^,[44]^11 Electroacupuncture at specific somatic tissues activates autonomic nervous system to induce either anti- or pro-inflammatory effects.[45]^12^,[46]^13^,[47]^14 Here, we revealed that thermal or chemical stimulation of TRPV1+ peripheral somatosensory nerves at nape could suppress inflammatory response via somato-autonomic reflex. Results TRPV1+ peripheral nerve stimulation induced the anti-inflammatory effects Nonivamide or pelargonic acid vanillylamide (PAVA), a less-pungent capsaicin analog, is a specific agonist of TRPV1 ion channel and used for anti-inflammation and relief of muscle pain in clinical application.[48]^15^,[49]^16 The toxicity of PAVA was evaluated by different administration routes. In our experiments, PAVA was administered at a dose of 4 mg/kg intraperitoneally (i.p.), 4 mg/kg subcutaneously (s.c.) into the nape (hereafter PAVA nape s.c.), 4 mg/kg s.c. into the abdomen (hereafter PAVA abdomen s.c.), 4 mg/kg s.c. into the back close to the tail (hereafter PAVA back s.c.), or 0.25 mg/kg intravenously (i.v.) into the tail veins ([50]Figure 1A). No signs of toxicity or death were found in the PAVA-treated mice. Previous studies have indicated that electroacupuncture stimulation at specific body regions (acupoints) can activate distinct neural circuits and modulate systemic inflammation in a somatotopy-dependent manner (body region specificity), while stimulation at non-acupoints fails to induce anti-inflammatory effects.[51]^12^,[52]^13^,[53]^14 Therefore, we designed a non-overlap layout scheme to select distinct injection positions on the body. PAVA was administrated via different routes to assess whether stimulating TRPV1+ afferents at specific body regions can modulate systemic inflammation induced by lipopolysaccharide (LPS), a bacterial endotoxin, through the specific neural pathway. LPS injected i.p. significantly increased serum levels of proinflammatory cytokines, including tumor necrosis factor alpha (TNF-α), interleukin (IL)-6, and IL-1β. Different PAVA treatments were performed 30 min before LPS challenge. Our PAVA treatments all substantially inhibited TNF-α at 1.5 h following LPS injection ([54]Figure 1B), when the release of TNF-α largely peaked in the serum. Notably, PAVA nape s.c. substantially reduced IL-6 at both 1.5 h and 6 h after LPS administration, whereas PAVA abdomen s.c. inhibited only IL-6 at 1.5 h, and the other treatments did not significantly reduce IL-6 ([55]Figure 1C). Serum IL-1β was barely detected at 1.5 h. Both PAVA nape s.c. and PAVA back s.c. caused a reduction in IL-1β at 6 h ([56]Figure S1A). Based on these results, PAVA nape s.c. was the most efficient anti-inflammatory method among the indicated treatments, suggesting that stimulation of TRPV1+ peripheral sensory afferents can inhibit cytokine production. PAVA nape s.c. was chosen for subsequent studies. Figure 1. [57]Figure 1 [58]Open in a new tab Stimulation of peripheral sensory afferents modulated inflammation (A) Schematic of distinct PAVA treatments and timeline of the experiments. (B and C) PAVA treatments in different body areas inhibited TNF-α (B) and IL-6 (C). (n = 4–6/group). (D and E) Distinct PAVA treatments at nape affected TNF-α (D) and IL-6 (E). (n = 4–6/group). (F and G) PAVA or dexamethasone treatment suppressed the release of TNF-α (F) and IL-6 (G). (n = 4–5/group). (H and I) β-alanine nape s.c. affected serum level of TNF-α (H) and IL-6 (I). (n = 4–5/group). Data are presented as the mean ± SEM. Two-way ANOVA with Bonferroni post tests, ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. Without LPS challenge, PAVA nape s.c. evoked baseline expression of TNF-α and IL-6 in serum ([59]Figures 1D and 1E). The anti-inflammatory effect of PAVA treatment was dose dependent. TNF-α level was reduced by 55.3% with 1 mg/kg PAVA and by 70.1% with 4 mg/kg PAVA ([60]Figure 1D). IL-6 was moderately but not significantly decreased with 1 mg/kg PAVA, while IL-6 was strongly reduced by 92% at 1.5 h and by 72.4% at 6 h with 4 mg/kg PAVA ([61]Figure 1E). Furthermore, CAP cream (PAVA 0.75 mg/g + capsaicin 0.25 mg/g) was transcutaneously (t.c.) applied to the shaved nape, which substantially reduced TNF-α and IL-6 ([62]Figures 1D and 1E), suggesting that stimulation of TRPV1+ cutaneous sensory afferents directly induce anti-inflammatory effects. Glucocorticoids are the most potent class of anti-inflammatory drugs and clinically used for treating various inflammatory diseases, but the severe side effects of glucocorticoid-treatments limit their long-term application, such as osteoporosis and insulin resistance. Dexamethasone, the synthetic glucocorticoid receptor agonist, was i.p. injected into the mice before LPS administration. PAVA nape s.c. inhibited the LPS-evoked release of TNF-α and IL-6 similar to dexamethasone pretreatment ([63]Figures 1F and 1G), suggesting a therapeutic potential of TRPV1+ peripheral afferent activation for treating inflammatory diseases. MRGPRD+ nociceptors are responsible for mechanical stimuli and are selectively activated by β-alanine to induce itch, and they constitute approximately 30% of total DRG neurons.[64]^8^,[65]^17^,[66]^18^,[67]^19 Single-cell RNA-seq studies indicated that MRGPRD was selectively expressed in NP1 subgroup and that TRPV1 was expressed in NP2, NP3, and PEP1 clusters in DRG.[68]^5^,[69]^6^,[70]^7 Pretreatment with β-alanine nape s.c. suppressed the LPS-evoked release of TNF-α but it increased IL-6 expression ([71]Figures 1H and 1I), indicating that stimulation of MRGPRD+ afferents induces a complex inflammatory response. Severe systemic inflammation is induced by the excessive release of a series of proinflammatory cytokines, and successful treatment is required to inhibit most of these cytokines rather than only a single cytokine. We examined the serum levels of 31 cytokines through a Luminex assay. Without LPS challenge, PAVA treatment significantly decreased the basal level of IL-6 and did not affect the expression of other cytokines ([72]Figures 2A and 2B). Luminex analysis revealed that LPS injection substantially elevated serum levels of 13 cytokines (fold > 1.5, p < 0.05) at 1.5 h, including TNF-α, IL-6, CCL2, CCL3, CCL4, CCL5, CCL7, CCL11, CCL12, CCL20, CCL22, CXCL1, and CXCL10. All of them were significantly inhibited by PAVA pretreatment, whereas the anti-inflammatory cytokine IL-10 was increased ([73]Figure 2A). Figure 2. [74]Figure 2 [75]Open in a new tab TRPV1+ peripheral somatosensory nerve stimulating protected the host against severe systemic inflammation (A and B) Heatmap showing serum cytokine levels at 1.5 h after distinct treatments. Fold change between LPS-only group and PAVA-LPS group (right). (n = 5/group). (C and D) With heat stimulation (∼50°C) or cold stimulation (∼0°C) of the shaved nape, heatmap showing serum cytokine levels at 1.5 h following LPS challenge. Fold change between LPS-only group and 50°C-LPS group (right). (n = 5/group). (E) PAVA nape s.c. increased the survival rate of endotoxemic mice (n = 25) compared to untreated endotoxemic littermates (n = 19). (F) PAVA nape s.c. improved the core body temperature of endotoxemic mice (n = 25) compared to untreated endotoxemic littermates (n = 19). (G) PAVA nape s.c. could not increase the survival rate of trpv1ko endotoxemic mice (n = 14) compared to untreated endotoxemic littermates (n = 13). (H) PAVA nape s.c. could not improve the core body temperature of trpv1ko endotoxemic mice (n = 17) compared to untreated endotoxemic littermates (n = 12). Data are presented as the mean ± SEM. Unpaired two-sided Student’s t test (A–D), Mantel-Cox log rank test (E and G), or two-way ANOVA with Bonferroni post tests (F and H). ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant. Capsaicin, a specific agonist of TRPV1 ion channel, can suppress inflammation via a TRPV1-independent mechanism by directly targeting PKM2-LDHA and COX2 in sepsis.[76]^20 To confirm whether TRPV1 ion channel is the actual target of PAVA in modulating its anti-inflammatory effect, we compared cytokine expression between wild-type and trpv1ko endotoxemic mice. Luminex analysis showed that the anti-inflammatory effect of PAVA treatment was completely reversed in trpv1ko mice ([77]Figures 2A and 2B). Furthermore, PAVA administration did not inhibit IL-1β expression in trpv1ko endotoxemic mice at 6 h following LPS challenge ([78]Figure S1B). Thus, PAVA treatment induced anti-inflammatory effects via TRPV1 ion channel. To investigate the contribution of TRPV1+ neurons and nerves in PAVA-induced anti-inflammatory response, mice were systemically treated with resiniferatoxin (RTX) to specifically induce TRPV1+ sensory neuropathy. Our previous studies indicated that TRPV1+ neurons in DRG can be divided into TRPV1^high and TRPV1^low subtypes according to the relatively high or low expression level of TRPV1.[79]^8^,[80]^21 The number of TRPV1^high neurons in DRG was significantly reduced in RTX-treated mice compared to control mice, while RTX treatment did not affect TRPV1^low neurons ([81]Figures S2A and S2B). TRPV1+ neuron ablation impaired the response to noxious heat ([82]Figure S2C). RTX-mediated sensory denervation significantly suppressed PAVA-induced anti-inflammatory effects on reducing the expression of TNF-α and IL-6 ([83]Figures S2D and S2E). Taken together, PAVA treatment induced anti-inflammatory effects through TRPV1+ sensory neurons and afferents. Thermal stimulation (>43°C) can elicit discharges of TRPV1+ cutaneous afferents. Heat stimulation (∼50°C) or cold stimulation (∼0°C) on the shaved nape was applied for 15 min ([84]Figure 2C). Luminex analysis showed that more cytokines were inhibited by heat stimulation than by PAVA treatment, while the administration of 4 mg/kg PAVA more effectively reduced the levels of some cytokines, such as IL-6 and CCL2, than heat treatment ([85]Figures 2A–2D). Moreover, heat stimulation promoted IL-10 expression in a manner similar to PAVA treatment. In contrast, only TNF-α and CCL3 levels were significantly decreased by cold stimulation ([86]Figures 2C and 2D). Therefore, stimulation of TRPV1+ cutaneous afferents at the nape with moxibustion-like treatment effectively suppressed proinflammatory cytokine production. Among the LPS-challenged mice, 60% of the mice pretreated with PAVA survived, while only 7% of the control littermates survived ([87]Figure 2E). The animals were monitored for 2 weeks, and no late deaths were found, indicating that PAVA treatment provides lasting protection. Moreover, PAVA treatment significantly normalized the core body temperature of endotoxemic mice at 12 h and 24 h following LPS challenge ([88]Figure 2F). In contrast, PAVA treatment did not improve the survival rate and core body temperature of trpv1ko endotoxemic mice ([89]Figures 2G and 2H). Taken together, these findings indicate that stimulation of TRPV1+ peripheral sensory afferents is an effective method for inhibiting systemic inflammation and protecting the host against lethal inflammation. Activation of NTS and C1 neurons in the brainstem via the somatosensory afferents To determine how PAVA nape s.c. transmit signals to the central nervous system, we used pERK expression to identify activated primary sensory neurons. PAVA nape s.c. significantly increased the number of pERK+ neurons in cervical (C3–C8) DRGs rather than thoracic (T3–T6) DRGs ([90]Figure S3A), indicating that the cervical somatosensory afferents were selectively activated. However, after PAVA administration, the number of pERK+ neurons in NGs was not increased compared to control ([91]Figure S3B). Therefore, neuronal signals activated by PAVA nape s.c. were mainly transmitted to the central nervous system via the cervical peripheral somatosensory afferents. To further explore which brainstem nucleus is excited following PAVA treatment, we scanned the expression of Fos, a marker of neuronal activation, in the hindbrain. PAVA nape s.c. strongly induced Fos labeling in NTS compared to control ([92]Figures 3A and 3B). In trpv1ko NTS, PAVA injection could not increase Fos+ neurons. NTS receives and integrates both vagal afferents and somatic afferents through spinal ascending pathway, and modulates vagal efferent nerves, and innervates the paraventricular nucleus (PVN) of the hypothalamus, which in turn activates the hypothalamic-pituitary-adrenocortical (HPA) pathway.[93]^22^,[94]^23^,[95]^24^,[96]^25 NTS plays an important role in controlling peripheral immune responses.[97]^26 In addition, PAVA administration strongly induced Fos expression in C1 neurons labeled with tyrosine hydroxylase (TH), but this Fos induction was eliminated in trpv1ko mice ([98]Figures 3C and 3D). C1 neurons located in the rostral ventrolateral medulla (RVLM) receive visceral and somatic sensory information and serve as a center to control multiple physiological functions. These cells modulate sympathetic efferent pathways.[99]^27^,[100]^28 Furthermore, these neurons also innervate PVN of the hypothalamus to modulate HPA axis. C1 neuron stimulation can activate both the splenic and adrenal sympathetic anti-inflammatory pathways.[101]^11^,[102]^29^,[103]^30 Thus, PAVA nape s.c. strongly activated NTS and C1 neurons in the medulla oblongata. Figure 3. [104]Figure 3 [105]Open in a new tab PAVA nape s.c. activated NTS and C1 neurons in the brainstem (A) PAVA nape s.c. induced Fos expression in NTS in wild-type mice but not in trpv1ko mice. Scale bar: 100 μm. (B) Quantification of Fos+ neurons in WT and trpv1ko NTS. (n = 3/group). (C) PAVA nape s.c. induced Fos expression (green) in TH+ (red) C1 neurons in wild-type mice but not in trpv1ko mice. Arrow, Fos+TH+ C1 neurons. Scale bar: 50 μm. (D) The percentage of Fos expression induced by PAVA treatment in TH + C1 neurons in WT and trpv1ko mice. (n = 3/group). Data are presented as the mean ± SEM, two-way ANOVA with Bonferroni post tests (B and D). ∗∗∗p < 0.001; ns, not significant. TRPV1+ peripheral afferent stimulating promoted corticosterone secretion Activation of hypothalamus promotes the pituitary gland to release adrenocorticotropic hormone (ACTH) into the bloodstream. Circulating ACTH drives the adrenal cortex to secrete glucocorticoids, which have multiple effects on immune cells to inhibit inflammation.[106]^31 To determine whether PAVA administration activates the HPA axis, we measured the expression of serum ACTH and corticosterone. Serum ACTH level was quickly increased at 5 min following PAVA treatment ([107]Figure 4A). Next, we performed a kinetic analysis of serum corticosterone expression, which indicated that the corticosterone level increased approximately 7-fold at 15 min following PAVA treatment and gradually decreased to baseline within 3 h ([108]Figure 4B). This PAVA-induced increase in serum corticosterone was eliminated in trpv1ko mice ([109]Figure 4C). To explore whether the induction of corticosterone modulates the protective effect of PAVA administration, we used metyrapone and mifepristone (RU486) to inhibit the synthesis of corticosterone and block glucocorticoid receptors, respectively.[110]^11^,[111]^32 Metyrapone pretreatment abolished the PAVA-evoked increase of corticosterone ([112]Figure 4D). However, neither treatment affected the anti-inflammatory effect of PAVA treatment on reducing the release of TNF-α and IL-6 ([113]Figures 4E and 4F). The possibility that HPA axis could play a redundant role or affect other immune responses was not ruled out. Figure 4. [114]Figure 4 [115]Open in a new tab PAVA nape s.c. increased serum level of ACTH and corticosterone (A) PAVA nape s.c. increased serum ACTH level at 5 min after stimulation. (n = 5/group). (B) Serum level of corticosterone at different time points following PAVA nape s.c. (n = 4–5/time point). (C) PAVA nape s.c. increased serum corticosterone release in wild-type but not trpv1ko mice. (n = 5/group). (D) Metyrapone pretreatment abolished the PAVA-induced increase of corticosterone. (n = 5/group). (E and F) Neither mifepristone (RU486) nor metyrapone pretreatment affected PAVA-evoked anti-inflammatory effects on the expression of TNF-α (E) and IL-6 (F) (n = 5/group). Data are presented as the mean ± SEM. Unpaired two-sided Student’s t test (A, E, and F), one-way ANOVA with Bonferroni post tests (B), or two-way ANOVA with Bonferroni post tests (C and D). ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant. The vagal-adrenal reflex increased circulating catecholamines Thermal cutaneous stimulation (>43°C) was reported to promote the secretion of adrenaline and noradrenaline via the autonomic-adrenal axis.[116]^33 Our kinetic examination of serum catecholamines following PAVA administration revealed increased expression. The levels of all of these molecules, predominantly adrenaline and dopamine, increased significantly at 40 min and 1.5 h after PAVA treatment, decreased to baseline at 3 h, and then increased again by 6 h ([117]Figures 5A–5C). This increase in PAVA-evoked serum catecholamines was eliminated in trpv1ko mice ([118]Figure 5D). Serum catecholamines are mainly produced by the chromaffin cells of the adrenal medulla, which are modulated by vagal and sympathetic preganglionic nerves through the release of acetylcholine.[119]^12^,[120]^13^,[121]^14^,[122]^34 Hexamethonium (HEX) is a nonselective nicotinic cholinergic receptor (nAChR) antagonist. Both subdiaphragmatic vagotomy (sVX) and HEX treatment completely abolished the induction of dopamine and noradrenaline ([123]Figures 5E and 5F), indicating that stimulation of TRPV1+ peripheral sensory afferents can drive the vagal-adrenal reflex to secrete serum catecholamines by signaling through nAChR. However, the induction of adrenaline was not effectively blocked by sVX or HEX treatment ([124]Figure 5G), suggesting that the induction of adrenaline is only partially dependent on the vagal-adrenal reflex. Some reports have indicated that immune cells can also generate and release catecholamines.[125]^35^,[126]^36 Adrenalectomy was performed to abolish the production of circulating catecholamines and corticosterone. Compared with their non-adrenalectomized littermates, the adrenalectomized mice exhibited substantially increased LPS-evoked cytokine production ([127]Figures 5H and 5I), indicating that serum catecholamines and corticosterone are important for the modulation of systemic inflammation. However, adrenalectomy did not affect the anti-inflammatory effects of PAVA treatment ([128]Figures 5H and 5I). Because both circulating catecholamines and autonomic nervous system play crucial roles in modulating multiple physiological functions, we measured vital signs following PAVA treatment. Systolic blood pressure and heart rate were decreased after PAVA nape s.c. injection, suggesting that stimulating TRPV1+ peripheral afferents can affect visceral functions via the somato-autonomic reflex ([129]Figure S4). Figure 5. [130]Figure 5 [131]Open in a new tab PAVA nape s.c. induced serum catecholamine release (A–C) Serum levels of dopamine (A), noradrenaline (B), and adrenaline (C) at different time points following PAVA nape s.c. (n = 4–5/time point). (D) PAVA nape s.c. increased serum catecholamine release in wild-type but not trpv1ko mice. (n = 4–5/group). (E–G) Both sVX and HEX pretreatment inhibited the release of dopamine (E) and noradrenaline (F) but not adrenaline (G). (n = 5/group). (H and I) Adrenalectomy (ADX) did not affect PAVA-evoked anti-inflammatory effects on the expression of TNF-α (H) and IL-6 (I). (n = 5/group). Data are presented as the mean ± SEM. One-way ANOVA with Bonferroni post tests (A–C) or unpaired two-sided Student’s t test (D–I). ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant. Local catecholamine secretion in the spleen was required for anti-inflammatory effects Both the sympathetic and vagal preganglionic nerves innervate the celiac ganglion (CG) and modulate the postganglionic splenic nerve through the release of acetylcholine.[132]^37^,[133]^38^,[134]^39 Activated splenic nerve terminals secrete noradrenaline in the spleen, which in turn inhibits proinflammatory cytokine production.[135]^11^,[136]^13^,[137]^29^,[138]^40 Our study showed that PAVA administration increased local noradrenaline release in the spleen ([139]Figure 6A). PAVA treatment failed to inhibit the production of TNF-α and IL-6 in splenectomized mice ([140]Figures 6B and 6C), indicating that the spleen plays a critical role in the anti-inflammatory pathway activated by PAVA treatment. Reserpine, a blocker of the vesicular monoamine transporter, was injected i.p. to deplete catecholamine stores in peripheral sympathetic nerve endings.[141]^41 The anti-inflammatory effect of PAVA treatment was eliminated after the blockade of local noradrenaline release in the spleen ([142]Figures 6B and 6C). To explore which adrenergic receptors may respond to the protective effects of noradrenaline, we tested the effects of the antagonists prazosin, RX821002, metoprolol and ICI118551 to block α1, α2, β1, and β2-adrenoceptors, respectively.[143]^35 Blockade of the α1- or β2-adrenoceptor abolished the inhibition of TNF-α production but did not affect the release of IL-6 ([144]Figures 6D and 6E). We speculated that distinct adrenergic receptors regulate the production of some cytokines in a redundant manner, in which the anti-inflammatory effects are not abrogated by blocking one of them. Thus, noradrenaline released from splenic nerve terminals mediated the anti-inflammatory effects of PAVA treatment. Figure 6. [145]Figure 6 [146]Open in a new tab Noradrenaline secretion in the spleen was required for PAVA-induced anti-inflammatory effects (A) PAVA nape s.c. promoted the release of splenic noradrenaline. (n = 5/group). (B and C) Serum levels of TNF-α (B) and IL-6 (C) in mice subjected to sham surgery (control), splenectomy or reserpine treatment before PAVA administration and LPS challenge. (n = 5/group). (D and E) Noradrenaline regulated serum TNF-α (D) and IL-6 (E) depending on distinct adrenoceptors. (n = 5/group). Data are presented as the mean ± SEM. Unpaired two-sided Student’s t test. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant. TRPV1+ peripheral nerve stimulating mediated gene expression in the spleen RNA-seq analysis was performed to characterize the transcriptional changes in the spleen triggered by distinct treatments, including saline control, PAVA alone, LPS alone or PAVA-LPS treatment ([147]Figure 7A). Functional pathway analysis revealed that differentially expressed genes (DEGs) were enriched in cytokine-cytokine receptor interaction, TNF, IL-17, and Toll-like receptor signaling pathways in each of the comparisons ([148]Figures 7B–7D). PAVA treatment restrained the expression of some proinflammatory genes, such as map2k3 and pik3r3, which play key roles in multiple signaling pathways to promote cytokine production ([149]Table S1). Consistent with our Luminex analysis of serum protein levels, the mRNA levels of a large group of proinflammatory cytokines were substantially increased in the spleen following LPS challenge, while PAVA pretreatment inhibited most of these cytokines and increased anti-inflammatory IL-10 level ([150]Figures 7E–7G). In addition, RNA-seq data showed that treatment with PAVA alone significantly altered gene expression in spleen. Compared to that in the control littermates, the downregulation of proinflammatory cytokines in the PAVA-treated mice was most notable ([151]Figures 7E–7G). PAVA treatment not only induced anti-inflammatory effects in inflamed states but also inhibited the immune response under normal physiological conditions. Figure 7. [152]Figure 7 [153]Open in a new tab RNA-seq analysis revealed that PAVA treatment induced significant alterations in gene expression in the spleen (A) The number of DEGs induced by distinct treatments. (B–D) KEGG analysis showing enriched pathways for comparing control group to PAVA only group (B), comparing control group to LPS only group (C), and comparing LPS only group to PAVA-LPS group (D). (E–G) Heatmap showing the RNA levels of splenic cytokines (E), chemokines (F), and interferon-related genes (G) at 1 h following indicated treatments. The scale in the heatmaps shows FPKM values transformed to log[2](FPKM) values for color scaling. Autonomic-immune reflex was activated to suppress inflammation To identify which division of the autonomic-immune axis mediates the anti-inflammatory effect of PAVA treatment, we performed sVX to block both the vagal-adrenal and vagal-CG-splenic axes. However, sVX did not affect the anti-inflammatory effects of PAVA treatment ([154]Figures 8A and 8B). Previous studies demonstrated that α7 nicotinic acetylcholine receptor (α7nAChR) was present in the murine CG and spleen, and played a crucial role in the cholinergic anti-inflammatory pathway induced by vagus nerve stimulation.[155]^42^,[156]^43^,[157]^44 We next investigated the effect of methyllycaconitine citrate (MLA), a selective α7nAChR antagonist.[158]^45 Although MLA pretreatment suppressed LPS-induced TNF-α production, it did not abolish the inhibition of IL-6 release ([159]Figure S5). Blocking α7nAChR is not enough to eliminate the anti-inflammatory effects of PAVA treatment. Spinal cord transection (SCT) at thoracic level 2 (T2) abrogates the sympathetic-splenic reflex.[160]^46 T2 SCT eliminated the inhibition of TNF-α but did not affect IL-6 expression ([161]Figures 8A and 8B), suggesting that the anti-inflammatory effects induced by PAVA treatment are partially dependent on the sympathetic efferent pathway originating from the supraspinal (medullary) reflex. HEX, a ganglionic blocker that inhibits nicotinic cholinergic ganglionic neurotransmission, inhibits both sympathetic and vagal preganglionic nerves to activate the splenic sympathetic nerve.[162]^47 In addition, this treatment suppressed the release of serum dopamine and noradrenaline induced by PAVA administration ([163]Figures 5E and 5F). Surprisingly, following HEX treatment, PAVA administration substantially increased the LPS-evoked release of TNF-α and IL-6 ([164]Figures 8C and 8D), suggesting that TRPV1+ peripheral afferent stimulation activates both the sympathetic and vagal efferent pathways to suppress inflammation. HEX pretreatment switched the effect of PAVA treatment from an anti-inflammatory effect to a proinflammatory effect. It suggested that stimulating TRPV1+ peripheral nerves at the nape induced a composite immune response that causes a strong autonomically mediated anti-inflammatory effect and a weaker proinflammatory action masked by the anti-inflammatory functions. Previous studies have indicated that serum adrenaline promotes the inflammatory cascade by activating α-adrenoceptors.[165]^35^,[166]^36 Our studies showed that HEX treatment did not block the increase in serum adrenaline induced by PAVA administration ([167]Figure 5G). After the mice were treated with HEX in combination with prazosin or RX821002, the LPS-evoked production of TNF-α and IL-6 decreased compared to that observed in the mice treated with LPS only, HEX-LPS, prazosin-LPS and RX821002-LPS. In addition, PAVA treatment regained the ability to suppress the release of TNF-α and IL-6 in the HEX-prazosin pretreated mice ([168]Figures 8E and 8F), suggesting that PAVA-induced secretion of serum adrenaline promotes inflammatory cytokine production by activating α-adrenoceptors. In contrast to the results in the HEX-treated mice, PAVA administration did not increase cytokine expression in the reserpine-treated or splenectomized mice ([169]Figures 6A and 6B), suggesting that serum dopamine and noradrenaline induced via the vagal-adrenal axis are also involved in inhibiting inflammation. Taken together, these findings indicate that stimulating TRPV1+ peripheral afferents activates both sympathetic-immune and parasympathetic-immune reflexes to synergistically suppress systemic inflammation. Figure 8. [170]Figure 8 [171]Open in a new tab Stimulation of TRPV1+ peripheral afferents activated the autonomic-splenic axis to suppress inflammatory cytokine production (A and B) Serum TNF-α (A) and IL-6 (B) in mice treated with sVX or T2 SCT before PAVA administration and LPS challenge. (n = 5–6/group). (C and D) Serum TNF-α (C) and IL-6 (D) were increased with HEX pretreatment before PAVA administration and LPS challenge. (n = 4–5/group). (E and F) Serum TNF-α (E) and IL-6 (F) in the mice treated with HEX+prazosin or HEX+RX821002 before PAVA administration and LPS challenge. (n = 4–5/group). Data are presented as the mean ± SEM. Unpaired two-sided Student’s t test. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant. Discussion Cumulative evidence supports the idea that the somatosensory nervous system not only identifies and transmits sensory information from the skin and internal tissues but also regulates peripheral immune responses.[172]^48^,[173]^49 The anti-inflammatory effects of peripheral somatosensory afferent stimulation are dependent on specific stimulation sites. Electric stimulation at the hindlimb ST36 acupoint activates dopamine-induced anti-inflammatory effects through vagal-adrenal axis.[174]^12^,[175]^14 Electroacupuncture at the abdomen ST25 acupoint elicits splenic nerve endings to release noradrenaline via spinal-splenic sympathetic reflex, which in turn suppress inflammation.[176]^13 Among our treatments, stimulating TRPV1+ peripheral nerves at the nape was the most effective method for suppressing cytokine production. It activated the HPA axis to rapidly secrete corticosterone, drove the vagal-adrenal axis to generate serum catecholamines, and provoked autonomic-splenic axis to suppress cytokine production. Through the somato-autonomic reflex, both the sympathetic and parasympathetic efferent pathways were activated to synergistically induce anti-inflammatory effects. In addition, TRPV1+ peripheral nerve stimulation at the nape activated NTS and C1 neurons in the brainstem. T2 SCT partially eliminated the anti-inflammatory effects. Taken together, these findings indicate that stimulating TRPV1+ peripheral nerves at the nape suppress systemic inflammation via the supraspinal (medullary) reflex in which the brain receives sensory inputs and drives sympathetic and parasympathetic efferent pathways to modulate physiological functions. Spleen-innervating primary sensory neurons are TRPV1+ nociceptors predominantly located in left T8–T13 DRGs ipsilateral to the spleen.[177]^50 In spinal cord, spleen-innervating neurons are consistent with the location of the left sympathetic preganglionic neurons at T4–T9 spinal levels.[178]^46 Somatosensory information from the trunk and limbs is transmitted to the same or a nearby segment spinal cord where the sympathetic preganglionic nerves are elicited to modulate visceral functions. This somato-spinal sympathetic reflex has a strong segmental organization.[179]^13^,[180]^51 Dietary capsaicin can activate TRPV1+ nociceptors in left T8–T13 DRGs to enhance the splenic germinal center response and humoral immunity.[181]^50 In the present study, it seems possible that stimulating TRPV1+ somatic afferents at the abdomen or the back excite TRPV1+ nociceptors in T8–T13 DRGs, which in turn activate a segmental spinal sympathetic reflex to suppress cytokine production in the spleen. Future studies will explore which anti-inflammatory circuits are activated by stimulating TRPV1+ somatic afferents at the abdomen and the back. A number studies demonstrated that electric stimulation of vagal nerves and the dorsal motor nuclei of the vagus can attenuate inflammation, indicating that vagus afferent and efferent nerves are critical for modulating innate immunity.[182]^10^,[183]^11^,[184]^52 Previous studies and our study showed that stimulating TRPV1+ visceral sensory afferents (PAVA i.p.) could inhibit TNF-α production ([185]Figure 1B), while ablation of TRPV1+ vagal sensory neurons in NG promoted inflammation during bacterial infection, suggesting that TRPV1+ visceral somatosensory afferents and vagal sensory afferents play crucial roles in controlling immune response.[186]^53^,[187]^54^,[188]^55^,[189]^56^,[190]^57 Recent study demonstrated that activation of TRPA1+ or CALCA+ vagal sensory neurons could elicit NTS to induce anti-inflammatory effects.[191]^26 Both TRPA1 and CALCA highly overlap with TRPV1 expression in NG.[192]^9 We propose that TRPV1+ vagal afferents can identify increased inflammatory signals via distinct inflammatory mediator receptors expressed on nerves and subsequently activate the brain-visceral interactions to protect organs from excessive inflammation. This vagal-autonomic reflex, including vagal-sympathetic and vagal-vagal reflexes, is critical for monitoring organismal inflammatory state and rebalancing internal immune response. Accordingly, we anticipate that disruption of this reflex will lead to out-of-control of immune response and induce a wide range of inflammatory and autoimmune diseases. Therefore, both somato-autonomic and vagal-autonomic reflexes are crucial for maintaining immune homeostasis and modulating immune responses. A comprehensive adaptable response of the host to excessive external stimuli via the neural-visceral reflex is crucial for stress resistance and survival. Here, we found that stimulation of TRPV1+ peripheral somatosensory afferents modulated immune responses and visceral functions, such as the secretion of serum corticosterone and catecholamines, blood pressure, and heart rate. Excessive external stimuli may affect internal homeostasis via somato-autonomic and somato-hormonal reflexes and neuroimmune communication. It is noteworthy that injury and pathogen infection in peripheral tissues excite nociceptors, pruriceptors, or thermoreceptors, which may transiently or continuously affect immunity and internal organ function. Our study revealed a specific neural circuit that stimulating TRPV1+ somatosensory afferents at the nape could concurrently drive both sympathetic and parasympathetic efferent pathways to synergistically induce anti-inflammatory effects. Stimulation intensity, somatosensory neuron specificity and body region specificity are all crucial for stimulating somatosensory fibers to modulate immune responses. Otherwise, treatment may cause complex or unwanted adverse reactions, such as stimulation of MRGPRD+ afferents, or may be insufficient to achieve the desired effects. Due to this endogenous mechanism, stimulation of TRPV1+ peripheral sensory afferents in specific body areas provides an effective strategy for preventing and treating inflammatory diseases. Limitations of the study Previous studies reported that male and female mice show differently sensitivity to LPS.[193]^13^,[194]^52 Therefore, this study focused on male mice. Female mice will be studied in future investigations to develop appropriate therapeutic approach for different gender. PAVA nape s.c. decreased systolic blood pressure and heart rate, which would be undesirable side effects in future clinical application. Although it seems possible that stimulating TRPV1+ peripheral afferents can affect visceral functions via the somato-autonomic reflex, this requires further investigation. Future studies are needed to explore whether appropriate stimulation intensity and specific stimulation position can treat systemic inflammation without affecting the blood pressure and heart rate. In addition, future studies are also needed to evaluate the therapeutic potential for TRPV1+ peripheral nerve activation in hypertensive patients. Resource availability Lead contact Further information and requests for resources and reagents should be directed to the lead contact, Zijing Liu (liuzijing@hotmail.com). Materials availability This study did not generate new unique reagents. Data and code availability * • RNA-seq data have been deposited at GEO and are publicly available as of the date of publication. Accession number is listed in the [195]key resources table. * • This paper does not report original code. * • Any additional information required to reanalyze the data reported in this paper is available from the [196]lead contact upon request. Acknowledgments