Kappa opioids suppress histaminergic itch by inhibiting mediodorsal thalamus-projecting claustrum glutamatergic neurons Authors: Zihan Liu, Xuelian Shu, Zhiwen Wang, Songyu Yao, Chuyuan Hu, Biying Wang, Panwen Liu, Lijin Feng, Yiquan Shen, Hui Xiao, Jing-Gen Liu, Yu-Jun Wang Published: 2026 ABSTRACT Itch, an uncomfortable sensation that induces scratching, is suppressed by activation of kappa opioid receptors (KORs). KOR agonists, such as Nalfurafine, offer potent central antipruritic effects, yet the underlying mechanisms are not fully understood. In this study, we report the important role of glutamatergic neurons in the claustrum (CLA) in itch modulation. Chemogenetic inhibition of these neurons significantly reduces histamine-induced scratching behavior, whereas their activation heightens scratching behavior. These CLAGlu neurons connect to the mediodorsal thalamus (MD), forming a pathway that modulates itch sensations. Optogenetic manipulation of the CLAGlu-MD pathway demonstrates that its inhibition decreases scratching behavior, whereas activation increases it. Notably, CLAGlu neurons highly express KORs, and treatment with Nalfurafine markedly inhibits histamine-induced CLAGlu neuronal activation. Conditional knockdown of KORs in CLAGlu neurons or the CLAGlu-MD circuit diminishes Nalfurafine’s antipruritic effects. These findings provide insight into the neural circuits involved in histaminergic itch and the action of KOR agonists. INTRODUCTION Itch, as a complex phenomenon and an uncomfortable sensory experience (1), frequently triggers scratching behaviors, leading to an itch-scratch cycle. The itch-scratch cycle can result in severe cutaneous injury (2, 3). Itch is broadly classified into two primary subtypes: histaminergic and non-histaminergic. Histaminergic itch is induced by histamine through the activation of histamine receptors, while nonhistaminergic itch results from other pruritogens, such as exogenous chloroquine and endogenous 5-hydroxytryptamine (5-HT), which activate non-histamine receptors (4, 5). Significant advancements have been made in understanding the peripheral and spinal coding of itch, including the identification of specific receptors, the neural pathways, and the molecular mechanisms involved in transmitting itch signals from the periphery to the spinal cord (6, 7). Functional brain imaging studies have revealed that the processing of itch and scratching activates broad supraspinal structures involved in the regulation of sensory, emotional, cognitive, motivational, and motor functions (8–11). Itch engages multiple areas of the brain, including the somatosensory cortex, the anterior cingulate cortex, the insula, and the prefrontal cortex (12). We previously found projections from the anterior cingulate cortex (ACC) to the dorsal medial striatum (DMS) modulated histaminergic itch-related behaviors and identified the ACC-DMS pathway as the downstream target of the anteromedial thalamic nucleus (AM) (13, 14). Despite these findings, the brain mechanisms that drive or modulate the itch-scratch cycle remain largely unexplored. The Claustrum (CLA), a vital brain hub involved in integrating sensory information from multiple modalities, is primarily composed of glutamatergic neurons (15). PET imaging studies on itch have found increased activity in the CLA and its connected regions during itch induction, suggesting its role in both sensation and modulation of itch (16, 17). The CLA receives inputs from a broad array of cortical areas, and sends extensive projections to various brain regions involved in processing itch sensation (15, 18–24). These studies point to the CLA as a potential key player in the neural circuitry of itch, but its specific contributions, related neural circuitry, and molecular mechanisms in itch modulation remain unknown. The CLA is known to highly express κ-opioid receptors (KORs) (25, 26). KORs, which are G protein-coupled receptors, play a critical role in modulating itch (27, 28). Numerous studies have demonstrated that activation of the KORs produces antipruritic effects (29–34). Clinically, Nalfurafine, a selective KOR agonist, has been utilized to alleviate pruritus in hemodialysis patients (35–37). The antipruritic effect of Nalfurafine is reported to be almost completely blocked by antagonizing central KORs (38), and further studies indicate that spinal cord KORs are the primary target where KOR signaling suppresses the transmission of itch inputs to supraspinal regions (39). However, the role of KORs in the CLA in itch modulation has not been explored, and the central mechanism by which KOR agonists like Nalfurafine modulate itch remains unknown. The involvement of CLA KORs in KOR opioid-induced antipruritic effects requires investigation. To address these questions, we dissect the function of CLA KORs and investigate the role of CLA and its related neural circuit in itch modulation. We show that CLA glutamatergic neurons (CLAGlu) modulate histaminergic itch through connecting to the mediodorsal thalamus (MD). Knockout of KORs in the CLAGlu neurons diminishes the anti-scratching effects of KOR agonists like Nalfurafine and U50488H. These findings advance our understanding of histaminergic itch circuits and KOR-mediated antipruritic mechanisms. RESULTS Multiple brain regions are active following histamine injection We intradermally injected mice with histamine (10 μg μl−1), which triggered significant scratching behaviors (Fig. 1, A and B). To identify the brain regions activated during scratching behavior, we quantified the expression of the immediate early gene c-fos (a marker of neuronal activity) in ten distinct brain areas (Fig. 1C). We observed that histamine administration significantly increased the number of c-fos-positive cells in several brain regions, including the prefrontal cortex (PFC), claustrum (CLA), paraventricular nucleus (PV), and parabrachial nucleus (PBN) (Fig. 1D). Multiple brain regions are active following histamine injection. (A) Schematic of histamine injection and experimental timeline. (B) Increased scratching bouts induced by intradermal injection of histamine in mice. ****P < 0.0001, Student’s t test, n = 5 to 6 mice. (C and D) Histamine-induced itch increased c-fos expression in multiple brain regions. (C) Representative images and (D) quantification of c-fos cells per mm2 in saline and histamine-treated mice. Scale bar, 100 μm. *P < 0.05, **P < 0.01, Student’s t test, n = 4 to 7 mice. All data are shown as mean ± SEM. PFC, prefrontal cortex; CLA, claustrum; PV, paraventricular nucleus; PBN, parabrachial nucleus. CLA glutamatergic neurons positively regulate histamine-induced scratching behaviors The CLA, given its extensive connectivity with sensory and limbic regions, has been proposed as a potential integrator in the perception of sensations (15, 40), including itch. It is predominantly composed of excitatory glutamatergic neurons (∼90%), with a small number of inhibitory interneurons (41). To identify the types of neurons activated by histamine in the CLA, we performed in situ hybridization to examine the colocalization of c-fos with the glutamatergic neuron marker (vGlut1 mRNA) or the GABAergic neuron marker (GAD65 mRNA). We found that 95.58% of the histamine-activated c-fos+ neurons in the CLA were colocalized with glutamatergic neurons, with only 3.22% localizing to GABAergic neurons. Moreover, after histamine injection, the proportion of glutamatergic neurons expressing c-fos increased from 15.70% to 34.42%, whereas no significant change was observed in c-fos-expressing GABAergic neurons (Fig. 2, A and B). To investigate the neural activities linked to itching, we employed fiber photometry to monitor intracellular Ca2+ levels. We injected AAV2/9-CaMKIIα-GCaMP6s into the CLA and implanted the optic fiber above the CLA (Fig. 2, C and D). We observed that the CLA glutamatergic neurons were significantly activated following histamine injection (Fig. 2, E to H). To verify the role of CLA glutamatergic neurons in itch sensation, we inhibited these neurons by bilaterally injecting a recombinant adeno-associated virus (rAAV) encoding hM4Di, an inhibitory designer receptor exclusively activated by a designed drug (DREADD), fused with mCherry under the control of CaMKIIα promoter (AAV2/9-CaMKIIα-hM4Di-mCherry) (Fig. 2I) into the CLA. The efficacy of hM4Di-mediated inhibition was confirmed by c-fos staining (Fig. 2, J and K). Intraperitoneal injection of the DREADD agonist clozapine N-oxide (CNO, 4 mg kg−1) significantly decreased histamine-induced scratching behavior (Fig. 2L). We next determined if chemogenetic activation of CLA glutamatergic neurons could promote itch. We injected AAV2/9-CaMKIIα-hM3Dq-mCherry into the CLA. Following CNO injection, there was a significant enhancement in c-fos expression and scratching behaviors induced by a low dose of histamine (0.25 μg μl−1) (Fig. 2, M to O). These data support the key role of CLA CaMKIIα neurons in histamine-induced sensation. CLA glutamatergic neurons are crucially involved in histamine-induced itch. (A and B) Histamine-induced c-fos expression within CLA colocalized with vGlut1 mRNA, but not with GAD65 mRNA. (A) Left: Representative images, including c-fos (white), vGlut1 (green) and DAPI (blue); Right: Quantitative analysis (per mm2). Scale bar, 50 μm. n = 3 mice. (B) Left: Representative images of the CLA, including c-fos (white), GAD65 (red) and DAPI (blue); Right: Quantitative analysis (per mm2). Scale bar, 50 μm. n = 3 mice. (C) Schematic diagram and experimental procedure. (D) Representative image. Scale bar, 200 μm. (E to H) Representative traces of Ca2+ signals of CLA glutamatergic neurons. (E) Synchronization of Ca2+ signals with markers of scratching behavior. (F) Representative traces of Ca2+ signals (red line) during scratching occur (−10 s − +10 s). n = 4 mice. (G) Heatmaps across animals aligned to the beginning of scratching. (H) Quantification of the area under the curve (AUC, −10–0 s, 0 s–10 s) of average ΔF/F. **P < 0.01. Student’s t test. n = 4 mice. (I) Schematic diagram and experimental procedure. (J and M) Representative images of c-fos cells among CLA glutamatergic neurons. Scale bar, 50 μm, 20 μm. (J) The inhibitory effects of the hM4Di. (M) The excitatory effects of the hM3Dq. (K and N) Quantification of c-fos and mCherry co-labeled cells. (K) **P < 0.01. Student’s t test, n = 3 mice. (N) ***P < 0.001. Student’s t test, n = 3 mice. (L) Chemogenetic inhibition of CLA glutamatergic neurons decreased histamine-induced scratching behavior. *P < 0.05. Student’s t test, n = 15 to 16 mice. (O) Chemogenetic activation of CLA glutamatergic neurons increased low-dose histamine-induced scratching behavior. *P < 0.05. Student’s t test, n = 8 to 9 mice. All data are shown as mean ± SEM. We further investigated the effects of other pruritogens, including chloroquine, 5-HT, and C48/80, on c-fos expression in the CLA. All these pruritogens triggered significant scratching behavior (fig. S1, A and B). In addition to histamine, chloroquine and C48/80 also increased c-fos expression in the CLA (fig. S1, C and D), whereas 5-HT had no effect. To rule out the possibility that increased c-fos expression in the CLA was due to scratching behavior itself, we examined c-fos expression following histamine injection into the plantar surface of the hindpaw (fig. S1E). We observed that plantar histamine injection similarly led to a significant increase in c-fos expression in the CLA (fig. S1, F and G). Pain and itch are two closely related but essentially distinct sensations that elicit different behavioral responses. To study the role of CLA glutamatergic neurons in nociceptive responses, we determined the effects of photoactivation of CLA glutamatergic neurons on pain responses induced by CFA (complete Freund’s adjuvant). We injected AAV2/9-CaMKIIα-ChrimsonR-mCherry and AAV2/9-CaMKIIα-GCaMP6s into the CLA, followed by the implantation of optical fiber over the CLA to enable selective activation (fig. S2, A to C). Optogenetic activation was confirmed using fiber photometry (fig. S2, D to G). We observed that optogenetic activation of unilateral CLA glutamatergic neurons not only inhibited CFA-induced allodynia but also increased paw withdrawal thresholds in the contralateral paw without CFA injection (fig. S2, H and I). These findings suggest that CLA glutamatergic neurons are involved in both pain and histamine-induced itch. CLA mediates KOR agonists Nalfurafine and U50488H-induced antipruritic effects To further investigate the causality of the CLA in itch sensation, we treated the mice with antipruritic agents, Nalfurafine and U50488H (Fig. 3A). Both Nalfurafine and U50488H significantly inhibited histamine-induced scratching behaviors (Fig. 3, B and C). Moreover, both Nalfurafine and U50488H treatment resulted in a significant reduction in the number of c-fos-positive neurons within the CLA (Fig. 3, D and E). To further verify the role of the CLA in mediating anti-itch effects of Nalfurafine and U50488H, we directly microinjected these compounds into the CLA (Fig. 3F). We observed that both Nalfurafine and U50488H inhibited histamine-induced scratching behaviors (Fig. 3, G and H). As both Nalfurafine and U50488H are selective KOR agonists, we microinjected the KOR antagonist nor-BNI into the CLA, followed by subcutaneous or intraperitoneal injections of Nalfurafine and U50488H (Fig. 3I). This treatment significantly suppressed the anti-scratching effects of both Nalfurafine and U50488H (Fig. 3, J and K). We also found that nor-BNI alone did not precipitate scratching behavior, and there was no difference between histamine and nor-BNI/histamine groups (fig. S3, A to C). Together, these data suggest that the CLA is a critical region mediating the central antipruritic effects of KOR agonists Nalfurafine and U50488H. Nalfurafine and U50488H exert antipruritic effects through the CLA. (A and B) Schematic diagram of experimental procedure. (C) Systemic injection of Nalfurafine and U50488H decreased histamine-induced scratching behavior. **P < 0.01, ****P < 0.0001. Student’s t test, n = 5 to 6 mice. (D and E) Nalfurafine and U50488H decreased histamine-induced c-fos expression in CLA. (D) Representative images and (E) quantification of c-fos cells in CLA. Scale bar, 50 μm. *P < 0.05, **P < 0.01, One-way ANOVA, n = 5 mice. (F and I) Schematic diagram of experimental procedure. (G) Microinjection of Nalfurafine and U50488H in the CLA decreased histamine-induced scratching behavior. **P < 0.01, ****P < 0.0001. Student’s t test, n = 6 to 10 mice. (H) Schematic representation of Nalfurafine and U50488H microinjection sites in the CLA. (J) Microinjection of nor-BNI into the CLA blocked the antipruritic effects of systemically administered Nalfurafine and U50488H. ***P < 0.001, ****P < 0.0001, One-way ANOVA, n = 8 to 12 mice. (K) Schematic representation of nor-BNI microinjection sites in the CLA. All data are shown as mean ± SEM. KORs in CLA glutamatergic neurons are crucial for Nalfurafine and U50488H-induced anti-scratching behaviors The CLA is known to express high levels of KORs and glutamatergic neurons, both of which play key roles in the regulation of itch. We then sought to determine whether KORs are expressed in CLA glutamatergic neurons. Results from in situ hybridization labeling of vGlut1, GAD65, and KOR mRNA demonstrated that 95.8% of KOR mRNA-positive neurons in the CLA co-localized with vGlut1 mRNA, while only 0.45% co-localized with GAD65 mRNA (Fig. 4, A and B). This indicates a significant overlap of vGlut1 and KOR mRNA in the CLA. Next, we investigated the role of KORs in CLA glutamatergic neurons in modulating Nalfurafine and U50488H-induced itch sensation. To examine this, we employed a knockout approach targeting KORs in CLA glutamatergic neurons and assessed the effects on Nalfurafine and U50488H-induced anti-scratching behaviors. We injected AAV2/9-CaMKIIα-Cre-EGFP into the CLA of Oprk1loxp/loxp mice (Fig. 4C), and KOR knockout efficacy was confirmed via in situ hybridization (Fig. 4, D and E). As shown in Fig. 4, F and G, the knockout of KORs in CLA glutamatergic neurons significantly decreased the anti-scratching effects of Nalfurafine and U50488H. Intra-CLA nor-BNI completely blocked the antipruritic effects of both Nalfurafine and U50488H (Fig. 3, I to K), whereas local KOR knockout in the CLA glutamatergic neurons only partially attenuated them (Fig. 4, C to G). This discrepancy suggests an important role of presynaptic KORs within the CLA. To test this, we used Oprk1loxp/loxp or wild type (WT) mice and injected either an anterograde AAV (AAV2/9-CaMKIIα-Cre-mCherry) or a retrograde AAV (AAV2/Retro-CaMKIIα-Cre-mCherry) into the CLA to selectively knock out KORs at different locations (Fig. 4H). Knockout efficiency was verified by in situ hybridization (Fig. 4, I and J). Furthermore, we validated the knockout efficiency in two known upstream brain regions of the CLA, the anterior cingulate cortex (ACC) and prelimbic cortex (PL) (fig. S4, A and B). We found that knocking out KORs both locally in the CLA and in presynaptic terminals from upstream projections suppresses the antipruritic effect of Nalfurafine to a greater extent than knocking out KORs only in local CLA neurons (Fig. 4K). Moreover, site-specific knockout of KORs did not alter spontaneous locomotor activity in mice (Fig. 4L). Together, these data support that presynaptic KORs on upstream terminals projecting to the CLA are a significant component mediating the antipruritic effects of Nalfurafine. Thus, both presynaptic KORs on terminals projecting to the CLA and KORs in CLA glutamatergic neurons are critical for mediating the antipruritic effects of Nalfurafine. The KORs in the CLA glutamatergic neurons are involved in the modulation of histamine-induced itch. (A and B) In situ hybridization of vGlut1 (green), GAD65 (red) and KOR (white) mRNAs reveals that the KOR mRNA primarily colocalizes with vGlut1 mRNA, but not GAD65 mRNA. (A) Representative images and (B) quantification of KOR mRNA colocalization with vGlut1 mRNA and GAD65 mRNA. Scale bar, 50 μm, 10 μm. (C) Schematic diagram of AAV injection into the CLA and experimental procedure. (D) In situ hybridization of KOR (white), vGlut1 (red), EGFP (green) and DAPI (blue) expression in the CLA after injection of AAV into the CLA. Scale bar, 50 μm, 20 μm. (E) Quantification of KOR colocalization with vGlut1 and EGFP. ****P < 0.0001. Student’s t test, n = 3 mice. (F and G) Knockout of KORs in CLA glutamatergic neurons decreased the antipruritic effects of Nalfurafine and U50488H. *P < 0.05. Student’s t test, n = 5 to 8 mice. (H) Schematic diagram of AAV injection into the CLA and experimental procedure. (I) In situ hybridization of KOR (white), vGlut1 (green), mCherry (red) and DAPI (blue) in the CLA after injection. Scale bar, 50 μm. (J) Quantification of KOR colocalization with vGlut1 and mCherry. ****P < 0.0001. One-way ANOVA, n = 3 mice. (K) Knockout of KORs both locally in the CLA and in presynaptic terminals from upstream projections suppresses the antipruritic effect of Nalfurafine to a greater extent than knocking out KORs only in local CLA neurons. **P < 0.01, ****P < 0.0001. One-way ANOVA, n = 9 to 11 mice. (L) Knockout of KOR did not alter spontaneous locomotor activity in mice. One-way ANOVA, n = 9 to 11 mice. All data are shown as mean ± SEM. The CLA-MD circuit critically mediates histamine-induced itch responses After identifying the participation of CLA glutamatergic neurons in histamine-induced itch and their mediation of KOR agonist-induced anti-scratching behaviors, we sought to elucidate the specific neuronal circuits involved in CLA regulation of itch. We injected AAV2/9-CaMKIIα-mCherry into the CLA, which resulted in robust labeling of the downstream region of the mediodorsal thalamic nucleus (MD) (Fig. 5A). For further confirmation, we injected ScAAV2/R-hSyn-mCherry, a viral-mediated retrograde tracer, into the MD of mice, and demonstrated the projections from the CLA to MD (Fig. 5B). We also found that the MD was significantly activated by histamine (Fig. 5, C to E). We further employed fiber photometry to monitor MD intracellular Ca2+ levels. We injected AAV2/9-CaMKIIα-GCaMP6s into the MD and implanted the optic fiber above the MD (Fig. 5F). We observed that the MD neurons were significantly activated following histamine injection (Fig. 5, G to I). Finally, to specifically record the CLAGlu-MD pathway during itch, we injected AAV2/9-CaMKIIα-Cre into the CLA and AAV2/R-DIO-GCaMP6s into the MD, and implanted an optic fiber above the MD (Fig. 5J). Following histamine injection, we observed that calcium transients in the MD-projecting CLAGlu axon terminals significantly increased during scratching episodes (Fig. 5, K to M). Interestingly, among the four pruritogens tested, only histamine increased c-fos expression in the MD, whereas chloroquine, 5-HT, and C48/80 did not (fig. S5, A to C). Similarly, histamine injection into the plantar surface of the hindpaw also induced elevated c-fos in the MD (fig. S5, D to F). MD is involved in histamine-induced itch. (A) Anterograde viral tracing from CLA to MD: schematic and representative fluorescence images. (B) Schematic and representative fluorescence images of retrograde viral tracing from CLA to MD. Scale bar, 100 μm. (C to E) Histamine-induced itch increased c-fos expression in the MD. (C) Schematic diagram of experimental procedure. (D) Representative images and (E) quantification of c-fos cells. Scale bar, 200 μm. *P < 0.05. Student’s t test, n = 8 mice. (F to I) MD glutamatergic neurons exhibit significant activation. (F) Schematic diagram and representative image and experimental procedure. Scale bar, 100 μm. (G to I) Representative traces of Ca2+ signals of MD glutamatergic neurons. (G) Representative traces of Ca2+ signals (blue line) during scratching occur (−10 s − +10 s). n = 3 mice. (H) Heatmaps across animals aligned to the beginning of scratching. (I) Quantification of the AUC (−10–0 s, 0 s–10 s) of average ΔF/F. ***P < 0.001. Student’s t test, n = 3 mice. (J to M) Axon terminals of CLAGlu neurons projecting to the MD exhibit significant activation during histamine-induced scratching behavior. (J) Schematic diagram and representative image and experimental procedure. Scale bar, 100 μm. (K) Representative traces of Ca2+ signals (brown line) during scratching occur (−10 s − +10 s). n = 3 mice. (L) Heatmaps across animals aligned to the beginning of scratching. (M) Quantification of the AUC (−10–0 s, 0 s–10 s) of average ΔF/F. ***P < 0.001. Student’s t test, n = 3 mice. All data are shown as mean ± SEM. Next, we focused on the CLA-MD circuit to determine its role in histamine-induced itch sensation. We used optogenetics to selectively manipulate MD-projecting CLA glutamatergic neurons to assess the causal role of the CLA-MD circuit in histamine-induced behaviors. We microinjected AAV2/9-CaMKIIα-ChrimsonR-mCherry or AAV2/9-CaMKIIα-eNpHR3.0-mCherry into the CLA, followed by the implantation of optical fibers over the MD to enable selective activation or inhibition (Fig. 6, A and B, F and G). Optogenetic modulation of the MD neural activity was confirmed using fiber photometry (Fig. 6, C and D, H and I). We observed that activation of the CLA-MD circuit increased low-dose histamine-induced scratching behaviors (Fig. 6E), while optogenetic inhibition of the CLA-MD circuit significantly inhibited histamine-induced scratching behaviors (Fig. 6, J to L). Furthermore, we found that photoactivation of the CLA-MD circuit also enhanced low-dose histamine induced scratching behaviors in female mice (fig. S6, A to F). Taken together, these findings suggest that the CLA-MD circuit mediates histamine-induced itch responses. CLAGlu-MD circuit regulates histamine-induced itch. (A) Up: Schematic diagram. Down: Experiment paradigm of light illumination and behavior test. (B) Representative images showing viral expression. Scale bar, 200 μm, 100 μm. (C and D) Representative traces of Ca2+ signals of the MD glutamatergic neurons. (C) The MD glutamatergic neurons in control mice (blue line) and photoactivation mice (red line) during the light close to open (−5 s − +10 s). n = 3 mice. (D) Heatmaps across animals aligned to the beginning of photoactivation. (E) Photoactivation of the CLAGlu-MD circuit increased low-dose histamine-induced scratching behavior. *P < 0.05. Student’s t test, n = 10 to 15 mice. (F) Up: Schematic diagram. Down: Experiment paradigm of light illumination and behavior test. (G) Representative images showing viral expression. Scale bar, 200 μm, 100 μm. (H and I) Representative traces of Ca2+ signals of MD glutamatergic neurons. (H) MD glutamatergic neurons in control mice (blue line) and photoinhibition mice (red line) during the light close to open (−5 s − +10 s). n = 3 mice. (I) Heatmaps across animals aligned to the beginning of photoinhibition. (J to L) Photoinhibition of the CLAGlu-MD circuit decreased histamine-induced scratching behavior. (J) *P < 0.05, **P < 0.01. Two-way ANOVA, n = 6 to 8 mice. (L) Paired comparison of scratching bouts during light-OFF and light-ON periods in eNpHR3.0-expressing mice. **P < 0.01, paired t test, n = 8 mice. All data are shown as mean ± SEM. The KORs in the CLAGlu-MD circuit mediate the antipruritic effects of Nalfurafine After establishing the critical role of the CLAGlu-MD pathway in histamine-induced itch, we next sought to determine whether the KOR agonist Nalfurafine exerts its effect through this circuit. We implanted a cannula into the CLA and injected AAV-CaMKIIα-GCaMP6s into the MD. After 4 weeks of viral expression, fiber photometry recordings were conducted during behavioral tests (Fig. 7, A and B). We found that intra-CLA administration of Nalfurafine reduced the activity of MD glutamatergic neurons compared to the saline group (Fig. 7, C to E). Subsequently, we further investigated the role of KORs in this circuit in Nalfurafine’s effect. We injected AAV2/9-CaMKIIα-fDIO-Cre-mCherry into the CLA and AAV2/R-CaMKIIα-FLP into the MD (Fig. 7F). We found that knockout of KORs in CLAGlu-MD decreased the anti-scratching effects of Nalfurafine (Fig. 7, G to I). However, this manipulation had no effect on CFA-induced mechanical allodynia (fig. S7, A to C). Together, these findings demonstrate that the KORs within the CLAGlu-MD circuit play an important role in the antipruritic effects of Nalfurafine. The KORs in the CLAGlu-MD circuit mediate histamine-induced itch. (A to E) Microinjection of Nalfurafine into the CLA suppresses MD neuron activity. (A) Schematic diagram and experimental procedure. (B) Representative image showing cannula implantation and viral expression. Scale bar, 200 μm, 50 μm. (C to E) Representative traces of Ca2+ signals of MD glutamatergic neurons. (C) The MD glutamatergic neurons in Saline mice (green line) and Nalfurafine mice (orange line) during pre- and post-injection (−150 s − +250 s). (D) Heatmaps across animals aligned to the beginning of injection. (E) Quantification of the AUC (0 s – 250 s) of average ΔF/F during the beginning of injection. *P < 0.05. Student’s t test, n = 3 mice. (F) Schematic diagram and experimental procedure. (G) Knockout of KORs in the CLAGlu-MD circuit decreased the antipruritic effects of Nalfurafine. P = 0.0556. Student’s t test, n = 10 to 13 mice. (H) In situ hybridization of KOR (white), vGlut1 (green), mCherry (red) and DAPI (blue) in the CLA. Scale bar, 50 μm, 20 μm. (I) Quantification of KOR colocalization with vGlut1 and mCherry. **P < 0.01. Student’s t test, n = 3 mice. All data are shown as mean ± SEM. DISCUSSION Understanding the brain regions and central circuits that underlie the itch-induced scratching behaviors is crucial, and recent research has made significant strides in this area (10, 42–45). In this study, we demonstrated that CLA glutamatergic neurons and the CLAGlu-MD circuit mediate histamine-induced itch. Importantly, CLA glutamatergic neurons and CLAGlu-MD projection express KORs, which is essential for mediating the antipruritic effects of the clinically used KOR agonist Nalfurafine (Fig. 8). Hypothetical model of glutamatergic neurons expressing KORs in the claustrum mediating histaminergic itch sensation and underlying the antipruritic effects of kappa opioids. Upon histamine injection, glutamatergic neurons in the claustrum (CLA) are activated, driving scratching behavior through their projections to the mediodorsal thalamus (MD). These CLA glutamatergic neurons exhibit high expression of KORs, which play a critical role in mediating the antipruritic effects of the central KOR agonist Nalfurafine. By activating KORs, Nalfurafine suppresses the activity of CLA glutamatergic neurons, thereby attenuating itch sensation and reducing pruritic responses. Created in BioRender. Ying, L. (2026) https://BioRender.com/rheafig. We identified a critical role for CLA glutamatergic neurons in histamine-induced itch. We observed increased CLA glutamatergic neuronal activity in response to histamine administration by c-fos staining and fiber photometry, and the activation of CLA glutamatergic neurons in turn facilitated itch signal processing, leading to the itch-scratching cycle. This notion is supported by experiments showing that the antipruritic agents Nalfurafine and U50488H significantly inhibit itch scratching behaviors and histamine-induced CLA activation. CLA densely expresses glutamatergic neurons (46, 47). Chemogenetic inhibition of CLA glutamatergic neurons results in a marked reduction in scratching behaviors, whereas activation of CLA glutamatergic neurons amplifies the scratching behaviors triggered by low-dose histamine. Our prior work demonstrates that modulation of CLA glutamatergic neurons did not affect general locomotor activity (46). Thus, these data indicated the critical role of CLA in itch sensation. CLA, a small nucleus between the insula and striatum, has been postulated to play a role in top-down control of cortical response to sensory input (48). Our results are consistent with previous studies, suggesting that CLA is critically involved in itch processing (9, 49, 50). Moreover, our observation that chemogenetic activation did not provoke spontaneous scratching but specifically increased low-dose histamine-induced scratching suggests CLA’s role as a modulator rather than an initiator of itch sensations, and further implies that CLA integrates input from various sensory modalities to intensify relevant stimuli (20, 48, 51). Thus, CLA acts as an integrative hub, facilitating the processing of multisensory information and refining the perception and behavioral response (20). It is important to note that while chemogenetic inhibition attenuated scratching, it did not reduce it to baseline control levels, and chemogenetic activation did not elevate scratching to the intensity seen with a full histamine challenge. This indicates that CLA glutamatergic neurons constitute a key modulatory node, but not the exclusive controller, for itch-related behavior, implying contributions from other brain regions. Indeed, our c-fos mapping studies showed that histamine activates a distributed network beyond the CLA, with significant increases in activity observed in the PFC, PV, and PBN. There was also a trend towards increased activity in the anterior cingulate cortex (ACC), intermediate lateral septum (LSI), and periaqueductal gray (PAG). These findings align with early positron emission tomography (PET) and functional MRI (fMRI) studies in humans, which have demonstrated the activation of multiple brain regions by pruritic stimuli (10, 16). Among these regions, previous studies have established the roles of the PAG, PBN, PFC, and ACC in modulating itch sensation (19, 52, 53), yet the function of the LSI in itch modulation remains to be further investigated. The mediodorsal nucleus (MD), a key component of the limbic thalamus, is characterized by dense interconnectivity within multiple brain regions, particularly limbic-associated structures (54). A recent study identified a thalamocortical circuit originating from the MD that is involved in scratching behaviors in response to both histaminergic- and non-histaminergic itch (55). In the present study, we identified that the MD was a major downstream target of CLA glutamatergic neurons. Notably, while multiple pruritogens, including histamine, chloroquine, 5-HT and C48/80, increased c-fos expression in the CLA, a significant increase in the MD was observed only following histamine injection, suggesting that the CLAGlu-MD circuit may play a distinct role in histamine-induced itch. Indeed, we further found that optogenetic inhibition of the CLAGlu-MD glutamatergic pathway significantly reduced histamine-induced scratching behaviors, while activation increased the itch response to low-dose histamine. Moreover, photoactivation of this pathway also enhanced itch response in female mice. A critical question is whether CLA glutamatergic neurons and the CLAGlu-MD pathway encode itch perception or merely reflect scratching-associated somatosensory feedback induced by intense mechanical stimulation. Our data support the former interpretation. Firstly, calcium activity in CLAGlu neurons begins to rise approximately 4 seconds prior to the onset of scratching bouts. Secondly, axonal calcium transients of CLAGlu neurons within the MD exhibit a similar pre-scratch increase. However, calcium signals in the broader MD neuronal population were directly aligned with the onset of scratching bouts. Thus, these data may suggest that the CLAGlu-MD pathway is functionally engaged in signaling itch perception, whereas the general population of MD neurons may primarily reflect scratching-induced sensory feedback. This is physiologically plausible, given that MD receives nociceptive and somatosensory inputs via direct spinothalamic and indirect parabrachial pathways (56–59). Recent studies have identified the claustrum (CLA) as a pivotal structure in the processing of nociceptive information and modulation of pain and itch sensations (60, 61). We found that optogenetic activation of glutamatergic neurons in the CLA increased mechanical thresholds in CFA-induced inflammatory pain, suggesting these neurons play dual roles: alleviating pain while exacerbating itch. This observation is consistent with studies highlighting the claustrum’s integral role in nociceptive processing (60, 61). Research by Faig and colleagues shows that selective activation of claustrocingulate neurons does not influence basic pain perception but does alleviate mechanical allodynia (61). Although the specific cell types remain unidentified, the dominance of glutamatergic neurons in the CLA (over 85%) suggests they might be crucial in mediating CLA’s contrasting effects on pain and itch. A finding was that CLA activation increased mechanical withdrawal thresholds in both the ipsilateral CFA-injected paw and the contralateral paw. CLA is reported to be a critical region in a cross-hemispheric inhibitory circuit (62). It is also well-established that activation of descending pain modulatory pathways (e.g., PAG-RVM-spinal cord) produces bilateral analgesia (63). The bilateral antinociceptive effect observed following unilateral CLA activation may reflect CLA engages cross-hemispheric inhibitory circuits or descending pain modulatory systems through its connections with ACC or other cortical regions, producing inhibition of pain processing bilaterally. KOR agonists are well-known for their antipruritic effects (64). Nalfurafine is a marked central KOR agonist, and U50488H, another KOR agonist, has demonstrated antipruritic properties (65). It was found that the inhibitory effect of TRK-820 (Nalfurafine) was completely antagonized by intracerebroventricular pretreatment with KOR antagonist nor-BNI, but not by subcutaneous administration (66). This suggests that the antipruritic effect of systemically administered Nalfurafine operates within the CNS. The central mechanisms of Nalfurafine mediating its anti-scratching behaviors are poorly understood. Recently, Sheahan with colleagues found that Nalfurafine relieves itch by selectively inhibiting GRPR spinoparabrachial neurons, demonstrating a mechanism where KOR signaling suppresses the transmission of itch inputs to supraspinal regions (39). Our findings identify the CLA and its projection to the MD as a critical supraspinal node mediating the effects of KOR agonists. Both Nalfurafine and U50488H inhibited histamine-induced neuronal activation in the CLA and MD. The CLA expresses high levels of KORs, predominantly in glutamatergic neurons. We established their functional significance by a knockout approach and observed that knockout of KOR expression either in CLA glutamatergic neurons or specifically in the CLAGlu-MD pathway significantly decreased the anti-scratching effects induced by Nalfurafine. It should be noted that we found that while local knockdown of KORs in the CLA only partially prevented the antipruritic effects of the KOR agonists, intra-CLA infusion of KOR antagonist nor-BNI achieved complete blockade. Moreover, we found that genetically ablating KORs both locally in the CLA and in presynaptic terminals of upstream projections suppressed Nalfurafine’s effect more robustly than ablating local KORs alone. These two lines of evidence support that the presynaptic KORs on afferent terminals projecting to the CLA constitute a key component mediating the antipruritic effects of Nalfurafine. Notably, ablation of KORs either locally within the CLA or including on presynaptic terminals of upstream projections did not affect spontaneous locomotor activity. This dissociation strongly supports a specific role for the CLA in itch modulation, rather than a general effect on motor function. MATERIALS AND METHODS Animals Male and female C57BL/6 mice weighing 18–22 g were purchased from Beijing Hua Fu Kang Biotechnology Co., Ltd. (Beijing, China). All mice were housed in a temperature- and humidity-controlled environment with free access to food and water under a 12-hour light-dark cycle (lights on at 07:00 and off at 19:00). The Oprk1loxp/loxp mice (Jackson number: 030076, USA) were used and genotyped according to the protocols provided by Jackson Laboratory. All animal experiments were performed under the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals using an approved animal protocol (2023–10-WYJ-05). Materials Nor-Binaltorphimine (nor-BNI, ab120078) was purchased from Abcam. (±)-U-50488 hydrochloride (U50488H, B6321) was purchased from APExBIO. Nalfurafine was purchased from Absin. Histamine (H7125) and the Complete Freund’s Adjuvant (CFA, F5881) were purchased from Sigma-Aldrich. Clozapine N-oxide (CNO, C08261) was purchased from Abcone. C-fos antibody (rabbit anti c-fos) (226003, RRID: AB_2231974) was purchased from Synaptic Systems Company. Secondary antibody (Alexa Fluor 488 goat anti-rabbit) (A-11008, RRID: AB_143165) was purchased from Thermo Fisher Scientific. AAV2/9-CaMKIIα-ChrimsonR-mCherry, AAV2/9-CaMKIIα-eNpHR3.0-mCherry, AAV2/9-CaMKIIα-mCherry, ScAAV2/R-hSyn-mCherry, AAV2/9-CaMKIIα-hM3Dq-mCherry, AAV2/9-CaMKIIα-hM4Di-mCherry, AAV2/9-CaMKIIα-Cre-EGFP, AAV2/9-CaMKIIα-EGFP, AAV2/9-CaMKIIα-Cre-mCherry, AAV2/R-CaMKIIα-Cre-mCherry, AAV2/9-CaMKIIα-Cre, AAV2/R-DIO-GCaMP6s, AAV2/R-CaMKIIα-FLP, AAV2/9-CaMKIIα-fDIO-Cre-mCherry, and AAV2/9-CaMKIIα-GCaMP6s were purchased from Brain VTA. All of the above viral titers were 2 × 1012–13 vg/mL, and all viral vectors were stored at −80°C. Itch test Mice were injected intradermally with 50 μL of histamine (10 μg μL−1) or 50 μL of histamine (0.25 μg μL−1) or 50 μL of chloroquine (4/0.8 μg μL−1) or 50 μL of 5-HT (1.5/0.3 μg μL−1) or 50 μL of C48/80 (2 μg μL−1) at the dorsal side of the neck. A digital video camera was used to record the itch-scratching behavior of mice over the subsequent 30 min. One bout of scratching was defined as a lifting of either hind paw to scratch the injection site and then returning it to the floor. Pain test In the Complete Freund’s Adjuvant (CFA)-induced inflammatory pain model, 20 μL of CFA was injected into the plantar area of the left hind paw. Mechanical pain thresholds of both hind paws were measured using von Frey filaments at 3 hours and 2 days after injection. During testing, mice were first placed in a chamber (transparent acrylic walls, wire mesh floor) for 15 min of acclimation, followed by 15 min of light activation. After activation, von Frey filaments (0.008 g, 0.02 g, 0.04 g, 0.07 g, 0.16 g, 0.4 g, 0.6 g, 1.0 g, and 1.4 g) were vertically applied to the midplantar area of both hind paws. Testing started with the 0.008 g filament, and each filament was applied five times. The pain threshold was recorded as the minimum filament force eliciting paw withdrawal (sudden retraction), flinching, or licking in three out of five trials. Cannula implantation and microinjection Mice were fixed on the stereotaxic apparatus (RWD Life Science) to expose the skull after being anesthetized with sodium pentobarbital (50 mg kg−1, i.p.). Afterwards, mice were implanted with cannulae after drilling holes in their skulls. The cannulae were bilaterally implanted above the CLA (AP: +1.18 mm; ML: ±2.5 mm; DV: −2.5 mm. AP, ML and DV denote the anteroposterior, mediolateral and dorsoventral distances from the bregma, respectively). Using stainless-steel screws and dental cement, the cannulae were anchored to the surface of the skull. Then, mice underwent behavioral testing seven days post-surgery. A 10 μL Hamilton microsyringe was connected to a 33-gauge needle and mounted on a microinfusion pump (Harvard Apparatus, USA). Bilateral microinjections of Nalfurafine (0.25 μL, 2 μg μL−1), U50488H (0.25 μL, 10 μg μL−1), or nor-BNI (0.25 μL, 10 μg μL−1) were performed using a needle extending 1 mm beyond the cannula. The infusion rate of all drugs was 0.1 μL/min. After infusion, the needle was left in place for 10 min to prevent the drug from leaking out. To determine the effect of intra-CLA microinjection of Nalfurafine on MD neuronal excitability, cannulas were bilaterally implanted above the CLA, and 200 nL of virus was microinjected into the MD at a rate of 0.1 μL/min. Tests were conducted 4 weeks after surgery. Bilateral microinjections of Nalfurafine (0.25 μL, 0.2 μg μL−1) were performed using a needle extending 1 mm beyond the cannula tip. After infusion, the needle was left in place for 10 min to prevent drug leakage. Viral injection and optical fiber implantation Mice were anesthetized intraperitoneally with sodium pentobarbital (50 mg kg−1, i.p.). For chemogenetic viral delivery, AAV2/9-CaMKIIα-hM3Dq-mCherry or AAV2/9-CaMKIIα-hM4Di-mCherry was microinjected into the CLA unilaterally or bilaterally (AP: +1.18 mm; ML: ±2.5 mm; DV: −3.5 mm) (0.15 μL/side) at a rate of 0.1 μL/min. After microinjection, the needle was placed for 10 min to ensure virus diffusion. Behavioral testing commenced four weeks post-viral expression. CNO (4 mg kg−1, i.p.) was injected 30 min pre-histamine to modulate CLA neuronal activity (excitation/inhibition). Mice were perfused 90 min post-CNO/saline injection, and c-fos immunofluorescence was quantified to validate manipulation efficacy. For optogenetic activation of the CLA, 0.15 μL of virus was microinjected into the unilateral CLA (AP: +1.18 mm; ML: ±2.5 mm; DV: −3.5 mm) at a rate of 0.1 μL/min. An optical fiber (diameter, 200 μm; 0.37 NA, Inper) was implanted above the CLA and secured to the skull by dental cement and stainless-steel screws. Mice were tested after four weeks of viral expression. For optogenetic activation of the CLA-MD pathway, 0.15 μL of virus was microinjected into the unilateral CLA (AP: +1.18 mm; ML: ±2.5 mm; DV: −3.5 mm), and 0.3 μL of virus was microinjected into the MD (AP: −1.34 mm; ML: 0 mm; DV: −3.6 mm) at a rate of 0.1 μL/min. An optical fiber (diameter, 200 μm; 0.37 NA, Inper) was implanted above the MD and secured to the skull by dental cement and stainless-steel screws. Mice were tested after four weeks of viral expression. For optogenetic inhibition of the CLA-MD pathway, 0.15 μL of virus was microinjected into the bilateral CLA (AP: +1.18 mm; ML: ±2.5 mm; DV: −3.5 mm) and 0.3 μL of virus was microinjected into the MD (AP: −1.34 mm; ML: 0 mm; DV: −3.6 mm) at a rate of 0.1 μL/min. An optical fiber (diameter, 200 μm; 0.37 NA, Inper) was implanted above the MD and secured to the skull by dental cement and stainless-steel screws. Mice were tested after four weeks of viral expression. Manipulation of KORs For knockout of KOR, 0.15 μL AAV2/9-CaMKIIα-Cre-EGFP was microinjected into the bilateral CLA at a rate of 0.1 μL/min. Mice were tested after four weeks of viral expression. Expressional validation of KOR was confirmed by in situ hybridization. To investigate the function of presynaptic KORs within the CLA, 0.15 μL AAV2/9-CaMKIIα-Cre-mCherry or AAV2/R-CaMKIIα-Cre-mCherry was microinjected into the bilateral CLA at a rate of 0.1 μL/min. Mice were tested after four weeks of viral expression. Expressional validation of KOR was confirmed by in situ hybridization. For knockout of KORs in the CLAGlu-MD, 0.15 μL AAV2/9-CaMKIIα-fDIO-Cre-mCherry was microinjected into the bilateral CLA and 0.2 μL of AAV2/R-CaMKIIα-FLP was microinjected into the bilateral MD at a rate of 0.1 μL/min. Mice were tested after four weeks of viral expression. Expressional validation of KORs was confirmed by in situ hybridization. CLA-MD disconnection To validate the reciprocal projections of the CLA-MD pathway, 0.1 μL AAV2/9-CaMKIIα-mCherry was microinjected into the CLA at a rate of 0.1 μL/min. Additionally, 0.2 μL ScAAV2/R-hSyn-mCherry was microinjected into the MD at a rate of 0.1 μL/min. Mice were sacrificed after viral expression, and the brains were removed and sectioned for fluorescence imaging under a confocal microscope. Optical manipulation The implanted fiber optics were connected to a fiber-coupled laser (adjustable power: 0–100 mW) through a rotary joint patch cable (1.25 mm OD, 0.37 NA, Inper). Optical stimulation was controlled by stimulus generator software via a computer. In the optogenetic activation test investigating CFA-induced pain, mice were injected with CFA (20 μL) and received 580 nm light stimulation (1 mW, 10 Hz, 5 ms pulses) for 10 cycles (60 s of stimulation and then rested for 40 s without light stimulation) after 3 hours and 48 hours. Mechanical hyperalgesia was assessed via the von Frey test to determine paw withdrawal thresholds after stimulation. In the optogenetic activation test investigating histamine-induced itch, mice were injected with histamine (50 μL, 0.25 μg μl−1) and immediately received 580 nm light stimulation (3 mW, 20 Hz, 5 ms pulses) for 6 cycles (3 min of stimulation and then rested for 2 min without light stimulation). Then, scratching bouts were recorded 30 min post-stimulation. For eNpHR3.0-mediated inhibition, mice were injected with histamine (50 μL, 10 μg μl−1) and immediately received 580 nm light stimulation (2 mW, 20 Hz, 5 ms pulses) for 6 cycles (5 min of stimulation and then rested for 5 min without light stimulation). At the same time, scratching bouts per 5 min of stimulation were recorded. Fiber photometry Fiber photometry was used to record calcium signals from glutamatergic neurons in the CLA, MD or CLAGlu-MD pathway using a three-color multichannel optical fiber recording system (Thinkertech). In recording changes in the CLA during scratching behavior, 0.15 μL of AAV2/9-CaMKIIα-GCaMP6s was injected into the CLA after mice were anesthetized, and an optical fiber (200 μm OD, 0.37 NA, Inper) was implanted into the CLA. After 4 weeks, mice underwent the scratching test, and Ca2+ fluorescence was recorded. Light at the fiber tip ranged from 10–20 μW (410 nm) and 20–40 μW (470 nm) and remained constant over trials during testing days. Change in fluorescence [ΔF/F = (F − F0)/F0] during scratching behavior (−10 s to 10 s) was analyzed using MATLAB. F0 was defined as the mean baseline signal from −2 s to 0 s. In recording changes in the MD during scratching behavior, 0.2 μL of AAV2/9-CaMKIIα-GCaMP6s was injected into the MD after mice were anesthetized, then an optical fiber (200 μm OD, 0.37 NA, Inper) was implanted. After 4 weeks, mice underwent the scratching test, and Ca2+ fluorescence was recorded. Light at the fiber tip ranged from 10–20 μW (410 nm) and 20–40 μW (470 nm) and remained constant over trials during testing days. Change in fluorescence [ΔF/F = (F − F0)/F0] during scratching behavior (−10 s to 10 s) was analyzed using MATLAB. F0 was defined as the mean baseline signal from −10 s to 0 s. In recording changes in the CLAGlu-MD pathway during scratching behavior, 0.15 μL of AAV2/9-CaMKIIα-Cre was injected into the CLA and 0.2 μL of AAV2/R-DIO-GCaMP6s was injected into the MD after mice were anesthetized, then an optical fiber (200 μm OD, 0.37 NA, Inper) was implanted. After 4 weeks, mice underwent the scratching test, and Ca2+ fluorescence was recorded. Light at the fiber tip ranged from 10–20 μW (410 nm) and 20–40 μW (470 nm) and remained constant over trials during testing days. Change in fluorescence [ΔF/F = (F − F0)/F0] during scratching behavior (−10 s to 10 s) was analyzed using MATLAB. F0 was defined as the mean baseline signal from −10 s to 0 s. To verify the effectiveness of light manipulating the CLA, 0.2 μL of a viral cocktail (1:1) containing AAV2/9-CaMKIIα-GCaMP6s and AAV2/9-CaMKIIα-ChrimsonR-mCherry/AAV2/9-CaMKIIα-mCherry was microinjected into the CLA after mice were anesthetized, then an optical fiber (200 μm OD, 0.37 NA, Inper) was implanted. After 4 weeks, mice underwent a von Frey test. Combining fiber photometry and optical manipulation, Ca2+ fluorescence was recorded when optical activation was achieved by 580 nm light (1 mW, 10 Hz, 5 ms pulses). Change in fluorescence [ΔF/F = (F − F0)/F0] during scratching behavior (−5 s to 15 s) was analyzed using MATLAB. F0 was defined as the mean baseline signal from −2 s to 0 s. To verify the effectiveness of light manipulating the CLAGlu-MD, 0.3 μL of AAV2/9-CaMKIIα-GCaMP6s was injected into the MD, and 0.15 μL of AAV2/9-CaMKIIα-ChrimsonR-mCherry or AAV2/9-CaMKIIα-eNpHR3.0-mCherry or AAV2/9-CaMKIIα-mCherry was injected into the CLA after mice were anesthetized. An optical fiber (200 μm OD, 0.37 NA, Inper) was implanted. After 4 weeks, mice underwent a scratching test. Combining fiber photometry and optical manipulation, Ca2+ fluorescence was recorded when optical activation or inhibition was achieved by 580 nm light (ChrimsonR: 3 mW, 20 Hz, 5 ms pulses; eNpHR3.0: 2 mW, 20 Hz, 5 ms pulses). Change in fluorescence [ΔF/F = (F − F0)/F0] during scratching behavior (−5 s to 10 s) was analyzed using MATLAB. F0 was defined as the mean baseline signal from −2 s to 0 s. Following recordings, mice were transcardially killed to confirm viral expression and optical fiber placement. If viral expression or optical fiber placement were incorrect, the behavioral data would not be analyzed. Immunohistochemistry Mice were deeply anesthetized with sodium pentobarbital (50 mg kg−1, i.p.) and perfused transcardially with saline, followed by 4% PFA in 0.1 M PBS, and the brains were removed. Mouse brains were first fixed with 4% PFA for 12 h, followed by 30% sucrose dehydration. Coronal sections (30 μm) were cut on a cryostat (Leica, Germany) after dehydration was complete. Then the coronal sections were stored in 0.1 M PBS. The sections were incubated in blocking buffer containing 3% normal goat serum in 0.5% Triton X-100/PBS for 1.5 h at room temperature and were incubated with primary antibodies — c-fos (1:1000) in blocking buffer overnight at 4°C. After four washes with PBS, sections were incubated with Alexa Fluor 488-conjugated secondary antibodies (1:1000) at room temperature for 2 h. After another four washes in PBS, sections were incubated with DAPI (Beyotime, C1005) for 30 min. Sections were dried and mounted, after which the fluorescence was observed using a confocal microscope. Quantitative analysis was performed with ImageJ software. In situ hybridization Mice were deeply anesthetized with sodium pentobarbital (50 mg kg−1, i.p.) and perfused with DEPC-PBS followed by 4% PFA in 0.1 M DEPC-PBS and their brains were removed. Brains were first fixed with DEPC-4% PFA for 12 h, followed by 30% DEPC-sucrose dehydration. Coronal sections (12–16 μm) were cut on a cryostat (Leica, Germany). Then, coronal sections were affixed to the slides (Citotest, China), dried, and stored at −80°C. The in situ hybridization procedures were performed according to the manufacturer’s instructions of RNAScope Multiplex Fluorescent Reagent Kit V2 (Advanced Cell Diagnostics, 323100). First, the slides were returned to room temperature and washed in DEPC-water, followed by 3 washes in 100% ethanol. Then, the slides were dried at 60°C for 2 h and cooled. Before boiling in target retrieval reagents for 7 min at 100°C, sections were incubated with hydrogen peroxide for 20 min. Next, the slides were rinsed with DEPC-water and 100% ethanol 3 times, respectively. Then, the sections were incubated with protease III for 30 min and incubated with probes for vGlut1-C1 (Slc17a7, ACD, 416631-C1), KOR-C2 (Oprk1, ACD, 316111-C2), GAD65-C3 (GAD65, ACD, 439371-C3), KOR-C1 (Oprk1, ACD, 316111-C1), vGlut1-C2 (Slc17a7, ACD, 416631-C2), EGFP-C3 (EGFP, ACD, 4000281-C3), mCherry-C3 (mCherry, ACD, 431201-C2), c-fos-C1 (c-fos, ACD, 316921-C1), GAD65-C2 (GAD65, ACD, 439371-C2) for 2 h after washing in DEPC-water 3 times. Subsequently, sections were incubated successively for Amplification step I (30 min), Amplification step II (30 min), and Amplification step III (15 min). Sections were then incubated with HRP-C1 for 15 min and then with C1-probe associated Opal 520 (1:750) for 30 min and subsequent incubation in HRP blocker solution for 15 min. Finally, sections were washed twice with a wash buffer between each incubation. The HRP-C2 and HRP-C3 procedures were the same as the HRP-C1, except for incubating with C2-probe associated Opal 570 (1:1000) and C3-probe associated Opal 690 (1:1500), respectively. At last, sections were incubated with DAPI for 5 min and treated with the Prolong Gold Antifade Reagent (Invitrogen, P36930) before being carefully mounted with a cover glass. All incubation steps were carried out at 40°C using a HybEz oven. Quantification and statistical analyses Sample sizes were determined based on previous experience. Animals were randomly assigned to experimental groups, and data analysis was performed blind to group allocation. All data are presented as mean ± SEM, with the number of biological replicates (n) specified in figure legends corresponding to independently treated experimental subjects. Data with mismatched fiber placement relative to the virus injection site were excluded. Statistical analyses were performed using GraphPad Prism, version 10.0, employing two-tailed unpaired Student’s t test or paired Student’s t test for single comparisons and one- or two-way ANOVA with Bonferroni post hoc comparisons for grouped analyses. Significance thresholds were set as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.