Chronic random sleep deprivation causes prolonged pain hypersensitivity via the anterior cingulate cortex

Summary

Sleep disturbance has a bidirectional relationship with pain. However, the effects of long-term sleep deprivation (SD) and its predictability on nociceptive behavior and neuronal activity remain largely unknown. In this study, we developed two mouse chronic SD models: a random model (RSD), in which mice could not predict when they would be able to sleep, and a fixed-schedule model (FSD). We investigated how these chronic SD patterns alter mechanical sensitivity. In both models, the mechanical threshold markedly decreased during SD, and neuronal activity in the anterior cingulate cortex (ACC) was facilitated. These behavioral and neuronal changes recovered in five weeks after the end of FSD, but persisted after RSD. Chemogenetic inhibition of ACC pyramidal neurons or administration of a chronic pain treatment, mirogabalin, normalized ACC activity, and attenuated the mechanical hypersensitivity in RSD mice. These results suggest that enhanced ACC activity mediates the SD-induced prolonged mechanical hypersensitivity.

Graphical abstract

Highlights

• •We developed a random sleep deprivation (RSD) model to assess its effects on pain
• •RSD mice exhibited prolonged mechanical hypersensitivity after sleep deprivation
• •Enhanced ACC activity correlated with RSD-induced mechanical hypersensitivity
• •Silencing ACC neurons reversed the prolonged mechanical hypersensitivity

Health sciences; Cell

Introduction

Good quality sleep is increasingly recognized as important for well-being, and sleep disturbance increases the risk of a range of diseases, including pain.¹ The circular bidirectional relationship between pain and sleep is well known in humans; pain and sleep deprivation (SD) aggravate each other to form a negative spiral of worsening impact.² Pain disrupts sleep,³ and consequently, patients with chronic pain have a higher incidence of sleep disturbance.^2,4 In addition, acute SD increases pain sensitivity,^5,6 and chronically restricted sleep has been reported to induce spontaneous pain and decreases tolerance to nociceptive stimuli in healthy individuals.⁷ Furthermore, patients with insomnia have a higher comorbidity of chronic pain.⁸ Sleep disturbance is also a risk factor for the development of pain chronicity after burn injury.⁹ However, the underlying mechanisms by which sleep disturbance induces and enhances pain hypersensitivity remain largely unknown.

Sleep duration and regularity are important factors affecting the quality of sleep. Several recent large human cohort studies have reported that sleep regularity can be a stronger predictor of harm than sleep duration for health outcomes^10,11,12 and mortality.¹³ How sleep deficiency and irregularity cause poor health consequences is largely unknown. Previous preclinical SD models have focused on sleep duration but not upon its predictability.^14,15,16,17 Therefore, there is a need for novel animal models to investigate the relationship between sleep regularity and adverse outcomes, such as pain chronicity, thereby identifying mechanisms and strategies to treat sleep-related adverse physical and mental health outcomes.

The anterior cingulate cortex (ACC) is important for pain processing, and plastic alterations in ACC synapses are thought to be one of the mechanisms underlying the development of chronic pain.^18,19 ACC activation produces pain hypersensitivity, and inhibition of ACC neuronal activity relieves chronic pain hypersensitivity.^20,21,22 Alterations in the ACC have been reported to mediate sleep disturbances induced by chronic pain,^23,24 and ACC activity is affected by chronic SD.²⁵ However, the relationship between ACC activity and pain hypersensitivity after chronic SD has not been investigated. Furthermore, the influence of the predictability of SD on pain phenotypes and ACC activity remains unexplored.

In this study, we developed two chronic SD models focusing on sleep duration and regularity (predictability): a random model (RSD), in which mice could not anticipate when they were to be allowed to sleep, and a fixed-schedule model (FSD). We assessed how these two chronic SD models alter nociceptive behavioral thresholds in mice and we tested whether ACC activity is involved in SD-induced pain alteration. We show that both regimes of sleep deprivation produce hyperalgesia but only RSD leads to prolonged hypersensitivity that lasts for at least 5 weeks after the end of sleep deprivation. ACC hyperactivity correlates with chronic SD-induced mechanical hypersensitivity during and after SD. Furthermore, inhibition of ACC neuronal activity or administration of the chronic pain treatment, mirogabalin, reversed SD-induced chronic mechanical hypersensitivity.

Results

Random chronic sleep disturbance produces prolonged mechanical and thermal hypersensitivity

We developed two chronic sleep disturbance models to investigate their effects on mechanical sensitivity. In the FSD model, which mimics chronic sleep disturbance present in scheduled shift workers, we kept the mice awake for the first 9 h of the light period (leaving 3 h undisturbed for sleep) and during the dark period by repeated tactile stimulation using a sweeper bar for 28 consecutive days (Figure 1A). In the RSD model, mimicking sleep disturbances present in unscheduled shift workers or patients with severe sleep disorders, mice were kept awake under the same conditions used for the FSD model, but the timing of the allowed sleep period (of 3 h) was randomly assigned in the 12-h light period and so was unpredictable for 28 consecutive days (Figure 1A).

Figure 1.

Both chronic sleep deprivation models caused prolonged mechanical and thermal hypersensitivity with different time courses

(A) Schematic of the sleep deprivation models, control (no sleep deprivation, top), a fixed-schedule sleep deprivation model (FSD, middle) and a random sleep deprivation model (RSD, bottom).

(B and C) Time course of changes in inactive periods of FSD and RSD mice before, during and after SD - during light phase with bar movement (top), light phase sleep period without bar movement (middle) and dark phase with bar movement (bottom). (FSD, n = 7; RSD, n = 8, two-way repeated measures ANOVA with Dunnet’s multiple comparisons test, ∗p < 0.05, ∗∗∗∗p < 0.0001, vs. pre; two-way repeated measures ANOVA with Bonferroni multiple comparisons test, †p < 0.05, ††††p < 0.0001, FSD vs. RSD) (C) Time course of the hindpaw mechanical withdrawal threshold of SD mice (control, n = 8; FSD, n = 12; RSD, n = 8, two-way repeated measures ANOVA with Tukey’s multiple comparisons test, ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, RSD vs. control group, $p < 0.05, $$$$p < 0.0001, FSD vs. control group, †p < 0.05, ††††p < 0.0001, FSD vs. RSD group).

(D) Latency to nocifensive behavior in hot-plate test before and 4 weeks after a 4 weeks exposure to RSD (n = 15, ∗∗∗∗p < 0.0001, Unpaired t test).

(E) Elevated plus maze and (F) open field tests both showing total distance traveled and time spent in open-arms/center zone, respectively (control (n = 11) vs. 4 weeks after exposure to RSD (n = 10), unpaired t test, ∗p < 0.05). Data summarized as mean ± SEM. The pale mauve bar indicates the timing of sleep deprivation (in B and C).

To analyze sleep pattern alteration in these SD models, we recorded motor activity (reflecting sleep state²⁶) using nano tag device²⁷ with a 2 min analysis window aligned to the bar movement schedule (once every 2 min) before and at two and four weeks after the start of SD. Two weeks after the start of SD, FSD, and RSD mice exhibited fewer periods of inactivity during the tactile stimulation periods in both the light and dark phases, consistent with decreased sleep duration as expected (Figure 1B, top and bottom). In the 3 h undisturbed window without tactile stimulation (sleep period), both FSD and RSD mice exhibited decreased periods of inactivity compared to before SD, suggesting that SD exposure altered sleep behavior even in the absence of stimulation (Figure 1B, middle). Four weeks after the start of SD exposure, FSD mice exhibited decreased periods of inactivity only during the dark phase, but not during the light phase SD period or the sleep period (Figure 1B). In contrast, RSD mice exhibited a decrease in the periods of inactivity in all phases (Figure 1B). This suggests that RSD and FSD exposure both disrupted the sleep patterns of the mice but in different manners.

We then performed pain-related behavioral tests in the chronic SD models to investigate the relationship to sleep disturbance and its predictability. As observed in other sleep disturbance models,^5,6 both FSD and RSD mice exhibited mechanical hypersensitivity during SD compared with the control group. Notably, RSD mice exhibited prolonged mechanical hypersensitivity after the end of SD, which lasted until the final time point that we measured (five weeks after the end of chronic SD, termed as RSD-hyperalgesic mice), whereas mechanical hypersensitivity in the FSD model mice gradually recovered after SD (Figure 1C). Furthermore, RSD mice exhibited prolonged heat hyperalgesia again lasting for weeks after the end of SD (Figure 1D). In the elevated plus maze (EPM) test and the open field test (OFT), the time spent in the open arm and time spent in the center zone were lower in RSD mice four weeks after the end of SD (Figures 1E and 1F), suggesting enhanced anxiety-like behavior. These results indicate that RSD disrupts sleep patterns, causes prolonged mechanical and heat hypersensitivity, and increases anxiety.

Activity of ACC neurons was enhanced in FSD and RSD mice, and the degree of facilitation correlated with mechanical hypersensitivity

To investigate the neural mechanisms underlying mechanical hypersensitivity induced by chronic SD, we focused on the ACC, which is involved in both mechanical hypersensitivity in chronic pain¹⁸ and SD-induced depressive behaviors.²⁵ To measure the activity of ACC pyramidal neurons, we performed multi-unit recordings under anesthesia and compared the mechanical pinch responses (for 10 s) between each group of SD mice. As reported previously, ACC neurons exhibit rhythmic firing under basal conditions (without stimulation),²⁸ and pinch stimulation induced a marked increase in the firing frequency of ACC neurons in all groups (Figures 2A–2D). While the basal firing frequency of ACC neurons were indistinguishable across the groups, the increase in firing frequency during and after pinch stimulation was enhanced in RSD and FSD mice four weeks after the start of chronic SD (Figures 2A–2D). Notably, the facilitation of ACC pinch-induced activity was also observed four to five weeks after the end of SD exposure in RSD mice, while responses in FSD mice were no different from control mice (Figures 3A–3D). Furthermore, the number of c-Fos positive cells in the ACC significantly increased in RSD mice four weeks after the end of SD compared to control mice (Figures 3E and 3F). These data suggest that the activity of ACC neurons was enhanced by both FSD and RSD exposure, and that the pattern of maintained enhancement in RSD corresponded with the prolonged mechanical hypersensitivity. In the following study, we focused on the prolonged mechanical hypersensitivity and enhanced neuronal responses exhibited in the mice four to five weeks after the end of RSD exposure.

Figure 2.

Pinch-evoked activity of ACC neurons is enhanced in FSD and RSD mice four weeks after the start of sleep deprivation

(A–C) Representative traces of ACC neuronal responses to pinch stimulation in control (A), FSD (B), and RSD mice (C) four weeks after the start of sleep deprivation. In these recordings, two to three units were isolated from the traces and their averaged waveforms (left) and unit activities are shown below each trace.

(D) Summary data of ACC neuronal responses (control, n = 24 cells, 4 mice; FSD, n = 25 cells, 3 mice; RSD, n = 37 cells, 5 mice, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test, ∗p < 0.05, ∗∗p < 0.01, FSD group vs. control group, †p < 0.05, ††††p < 0.0001 RSD group vs. the control group). Data are the mean ± SEM.

Figure 3.

Pinch-evoked activity of ACC neurons is still enhanced in RSD mice four weeks after the end of sleep deprivation

(A–C) Representative traces of ACC neuronal responses to pinch stimulation in control (A), FSD (B), and RSD mice (C) four weeks after the end of sleep deprivation. In these recordings, two to three units were isolated from the traces and their averaged waveforms (left) and unit activities are shown below each trace.

(D) Summary data of ACC neuronal responses (control, n = 13 cells, 3 mice; FSD, n = 25 cells, 3 mice; RSD, n = 24 cells, 4 mice, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test, ∗∗∗p < 0.001 RSD group vs. control group).

(E) Representative images of c-Fos staining in the ACC of control and RSD model mouse.

(F) Summary data of c-Fos positive cell counts in the ACC (n = 4 mice per group, unpaired t test, ∗p < 0.05). Data are the mean ± SEM.

Chemogenetic inhibition of ACC neurons alleviated the prolonged mechanical hypersensitivity in RSD mice

Next, we addressed whether the enhanced activity of ACC neurons is involved in mediating mechanical hypersensitivity in RSD-hyperalgesic mice using a chemogenetic approach. To inhibit the neuronal activity of ACC neurons, we expressed hM4Di, inhibitory designer receptor exclusively activated by designer drugs (DREADD),²⁹ bilaterally in ACC pyramidal neurons by injecting an adeno-associated virus (AAV), AAV_DJ-CaMKIIα-hM4Di-mCherry (Figure 4A). We observed a robust expression of mCherry (fusion with hM4Di) in the ACC bilaterally (Figure 4A). Slice electrophysiological recordings showed that CNO application (10 μM) decreased the resting membrane potential of mCherry-positive pyramidal neurons (Figure 4B, degree of hyperpolarization; −6.21 ± 0.93 mV) and also decreased the action potential frequency induced by depolarizing current pulses (Figures 4C and 4D). In vivo electrophysiological recording also showed that CNO (3 mg/kg, i.p) decreased the basal firing frequency and ACC neuronal responses to pinch stimulation in RSD-hyperalgesic mice (Figures 4E and 4F). Behavioral analysis showed that the decreased mechanical threshold in RSD-hyperalgesic mice was significantly ameliorated by ACC chemogenetic inhibition (Figures 5A and 5B), whereas similar CNO administration (3 mg/kg, i.p) did not affect the mechanical hypersensitivity in RSD-hyperalgesic mice transduced with a control vector (AAV_DJ-CaMKIIα-mCherry) (Figure 5C). These findings suggest that enhanced ACC pyramidal cell excitability is associated with the prolonged mechanical hypersensitivity in RSD-hyperalgesic mice.

Figure 4.

Chemogenetic inhibition of ACC neurons reduces pinch-induced neuronal activation

(A) Schematic of AAV (AAV_DJ-CaMKIIα-hM4Di-mCherry) vector injection into the bilateral ACC (left). A representative image showing mCherry (red)-expressing neurons in the ACC (middle and right—with Neurotrace).

(B) A representative trace of hyperpolarization induced by CNO application (left) and the summary data (right, n = 4 neurons, two tailed paired t test, ∗∗p < 0.01).

(C) Representative traces of AP firing response of the mCherry-positive ACC neuron induced by current injection before (Pre) and after CNO application (CNO, 10 μM).

(D) Summary data of AP responses of mCherry positive neurons before and after CNO (n = 6, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001).

(E) Representative traces of ACC neuronal responses to pinch stimulation pre-application (control, top), 30 min after (middle) and 60 min after CNO (3 mg/kg, i.p) (bottom) of RSD-hyperalgesic mouse made 4 weeks after the end of sleep deprivation.

(F) Summary data of pinch stimulation-evoked ACC neuronal responses normalized with respect to basal firing levels (upper: responses during pinch stimulation, below: activity in the 10 s after pinch stimulation, n = 9 cells, 3 mice, one-way repeated measures ANOVA with Dunnett’s multiple comparisons test, ∗∗∗∗p < 0.0001). Data are mean ± SEM.

Figure 5.

Chemogenetic inhibition of ACC neurons attenuates the mechanical hypersensitivity in RSD-hyperalgesic mice

(A) Schematic of AAV vector injection into the ACC bilaterally.

(B and C) Effects of CNO application on mechanical hypersensitivity in RSD-hyperalgesic mice expressing hM4Di-mCherry (B, CNO 3 mg/kg, n = 10; saline, n = 8, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test, ∗∗∗∗p < 0.0001) or RSD-hyperalgesic mice with control vector (AAV_DJ-CaMKIIα-mCherry) injection (C, CNO, n = 9 each, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test). Data are the mean ± SEM.

An anti-neuropathic pain treatment, mirogabalin, attenuates enhanced ACC activation, and mechanical hypersensitivity in RSD-hyperalgesic mice

Finally, we addressed whether prolonged mechanical hypersensitivity in RSD mice was ameliorated by the gabapentinoid, mirogabalin.³⁰ Gabapentinoid drugs bind to α₂δ-1 calcium channel subunits which are distributed in the CNS and PNS, to exert beneficial therapeutic effects on chronic pain.³¹ In vivo extracellular recording showed that administration of mirogabalin (3 mg/kg, i.p) reduced the enhanced ACC neuronal responses during and after pinch stimulation in RSD-hyperalgesic mice (Figures 6A and 6B). Furthermore, mirogabalin administration (3 mg/kg, i.p) also significantly attenuated mechanical hypersensitivity in RSD mice at the end of SD (Figure 6D) and four weeks after the end of SD (Figure 6E), while the mechanical sensitivity of control (naive) mice was not affected by mirogabalin (Figure 6C). In contrast, administration of the NSAID analgesic diclofenac (10 mg/kg i.p) had no effect on mechanical hypersensitivity in RSD-hyperalgesic mice (Figure 6F). These data suggest that gabapentinoid drugs could be effective medication for SD-induced hyperalgesia. To investigate the mechanisms of mirogabalin-induced analgesia, we injected mirogabalin directly into the ACC using a dual infusion cannula (Figures S1A and S1B). Mirogabalin infusion did not alter mechanical hypersensitivity in RSD-hyperalgesic mice, while muscimol, GABA_A receptor agonist, significantly ameliorated the mechanical hypersensitivity of RSD-hyperalgesic mice (Figure S1C). These data suggest that mirogabalin does not directly affects ACC activity but could exert its analgesic effect through an action at α₂δ-1 subunits in the other pain-related central and/or peripheral nervous system sites.

Figure 6.

Mirogabalin administration restored pinch-induced neuronal activation and the mechanical hypersensitivity in RSD-hyperalgesic mice

(A) Representative traces showing ACC neuronal responses to pinch stimulation in control saline-treated RSD-hyperalgesic mice (top) and mirogabalin-treated RSD-hyperalgesic mice (bottom) 4 h after administrations.

(B) Summary data showing effects of mirogabalin application on ACC neuronal responses normalized with basal firing rate (upper: responses during pinch stimulation, below: responses after pinch stimulation for 10 s; saline n = 13 cells, 3 mice; mirogabalin n = 9 cells, 3 mice, one-way repeated measures ANOVA with Dunnett’s multiple comparisons test, ∗p < 0.05, ∗∗∗p < 0.001).

(C) Effect of mirogabalin application on mechanical threshold of control (non-sleep derived) mice (n = 6 each).

(D) Effect of mirogabalin application on mechanical hypersensitivity exhibited in RSD mice at the end of SD exposure (n = 10 each, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test, ∗∗p < 0.01, ∗∗∗∗p < 0.0001).

(E) Effect of mirogabalin application on mechanical hypersensitivity exhibited in RSD mice 4 w after the end of SD (mirogabalin, n = 8; saline, n = 7, two-way repeated measures ANOVA with Bonferroni’s multiple comparisons test, ∗∗p < 0.01, ∗∗∗∗p < 0.0001).

(F) Effect of diclofenac application on mechanical hypersensitivity exhibited in RSD mice 4 w after the end of SD (n = 10). Data are the mean ± SEM. See also Figure S1.

Discussion

Sleep disturbance affects the activity of the whole brain, including the limbic regions,²⁵ and is associated with many problems, including mental health disorders, cardiovascular disease, metabolic disorders,^32,33 cognitive deficits,³⁴ and pain hypersensitivity.^2,7 While it is clinically well known that sleep deficiency and pain are linked and aggravate each other, the underlying brain mechanisms of SD-induced pain hypersensitivity are largely unknown.² Sleep duration and regularity are important for the quality of sleep, health conditions,^10,11,12 and mortality risk.¹³ Previous animal models used for chronic SD only focused on the sleep duration by limiting animal sleep opportunities or forced activity,^14,15,16,17 but did not focus on sleep regularity. In this study, to investigate the relationship between sleep irregularity and its consequences, we developed novel chronic SD models where in which sleep opportunity was chronically restricted to 3 h in a day, regularly (FSD) or irregularly (randomly, RSD) for four weeks. In the FSD protocol, the time spent inactive (reflecting sleep) was significantly reduced in both the light and dark phase two weeks after the start of SD, but this reduction only persisted in the dark phase four weeks after the start. In contrast, RSD exposure significantly reduced periods of inactivity in all phases of the light dark cycle both two and four weeks after the start of SD. Using these models, we showed that the mechanical hypersensitivity in RSD mice was prolonged beyond the time of SD exposure. We also noted that ACC hyperactivation was associated temporally with this hypersensitivity and chemogenetic inhibition of the pyramidal neurons ameliorated the hypersensitivity. Furthermore, systemic mirogabalin, a treatment for chronic neuropathic pain, showed a similar ability to improve the hypersensitivity induced by RSD.

Regular sleep habit is important for health and well-being, and night shift workers frequently complain that their sleep length and quality are altered by the variable retiring and rising times, and suffer from subsequent disturbances of psycho-physiological functions.³⁵ Our data show that RSD mice exhibited prolonged mechanical hypersensitization compared with FSD mice, but the underlying mechanisms by which unpredictable sleep disruption causes prolonged mechanical hypersensitivity are unclear. Previous studies have showed that irregularity in chronic sleep routines are related to behavioral and psychological deficits³⁶ and rotating night shift work is associated with higher risk of heart disease and breast cancer.^37,38 Therefore, irregularity of sleep schedules could aggravate the chronic SD-induced mechanical hypersensitivity in RSD mice. It is also unclear whether the sleep alteration and/or other physiological disfunctions induced by RSD would continue after period of sleep intervention, leading to the prolonged pain hypersensitivity. Further studies are needed to identify detailed psycho-physiological alterations by RSD exposure.

The ACC participates in sensory integration and processing of noxious and innocuous sensory stimuli, and is involved in mediating pain hypersensitivity,^18,39 but the relationship between alteration of ACC activity and SD-induced pain hypersensitivity has not been thoroughly investigated. A previous study showed that chronic SD increased the expression of c-Fos, a neuronal activity marker, in the isocortex-cerebral cortex plate area, including the ACC, suggesting that this enhanced activity is involved in anxiety behavior after chronic SD.²⁵ Furthermore, sleep deprivation causes pain hypersensitivity in human healthy volunteers,^40,41 and interestingly a human functional MRI study showed that pain-associated activations were amplified within primary somatosensory cortex and caudal ACC.⁴² Consistent with these reports, we found that the activity of ACC neurons was facilitated by chronic SD both during and after the SD period in RSD mice, which led to subsequent prolonged pain hypersensitivity. However, the mechanism underlying this hyperactivation of ACC neurons remains unknown. Dopamine signaling dysfunction was proposed to be the mechanism underlying both acute and chronic SD-induced pain hypersensitivity.⁶ A recent study reported that loss of dopamine signaling in the ACC induces hyperactivity of ACC pyramidal neurons and mechanical hypersensitivity in a mouse chronic pain model.⁴³ Thus, it is possible that decreased dopamine levels after chronic SD could induce ACC hyperactivation and mechanical hypersensitivity.

We also found that mechanical pain hypersensitivity was prolonged in both the RSD and FSD models, and the hypersensitivity lasted longer in the RSD model, suggesting that the predictability of SD was associated with the duration of prolonged pain hypersensitivity. Plastic changes in ACC neuronal synapses cause long-term mechanical hypersensitivity in mouse chronic pain models,¹⁸ therefore, synaptic alterations could be involved in prolonging pain hypersensitivity after chronic SD, and the strength of this nociplastic change can be altered by the type of SD models. Further studies are required to elucidate the underlying mechanisms.

Treatment options for SD-induced pain hypersensitivity are limited.¹⁸ A previous study showed that SD-induced pain hypersensitivity was not relieved by the analgesics, ibuprofen and morphine, but was ameliorated by caffeine and modafinil, two wake-promoting agents.⁶ We also found that administration of diclofenac, another NSAID, did not alter mechanical hypersensitivity in RSD-hyperalgesic mice. However, our results showed that a gabapentinoid, mirogabalin, reduced both pain hypersensitivity and hyperactivity in ACC neurons after RSD exposure. Gabapentinoids relieve neuropathic pain³¹ through binding to α₂δ-1 subunits, which are broadly expressed in the CNS^44,45 and PNS and are important for the function of voltage-gated calcium channels. In our experiment, intra-ACC infusion of mirogabalin did not ameliorate RSD-induced mechanical hypersensitivity, suggesting that mirogabalin exerts its analgesic effect by acting on α₂δ-1 subunits expressed in other CNS or PNS areas and could indirectly inhibit ACC hyperactivity. Pregabalin, a gabapentinoid, can also relieve widespread pain induced by direct chemogenetic central amygdala activation,⁴⁶ a type of CNS-derived nociplastic pain without nociceptor activation or nerve injury.⁴⁷ Pregabalin has also been reported to relieve sleep disturbance and pain symptoms in patients with fibromyalgia⁴⁸ which may be relevant to our mouse model findings. Because gabapentinoids are used as medication for generalized anxiety disorder,⁴⁹ anxiolytic effects of gabapentinoids could also beneficial for treatment for SD-induced anxiety which was noted in our model after RSD. Although we should consider the balance between the clinical effects of gabapentinoids and their side effects, including sedation and dizziness, our findings suggest that mirogabalin and perhaps other gabapentinoid drugs could be beneficial for restoring pain sensitivity after SD. Because we have not checked the effects of other analgesics on both pain hypersensitivity and enhanced ACC neuronal activity in RSD-hyperalgesic mice, they are important subject to investigate in the future study.

The relationship between pain and sleep is clinically well established; insufficient sleep may worsen pain, and patients with chronic pain often complain of sleep deficiency.² We found that ACC hyperactivity was associated with prolonged mechanical hypersensitivity after chronic SD. The ACC is important for pain processing¹⁸ and its activity is affected by sleep disorders and chronic pain.^25,39 Therefore, ACC may be a crucial link between sleep deficiency and pain. Our findings suggest that attenuation of ACC neuronal activity may be a useful strategy for normalizing pain hypersensitivity after chronic sleep problems.

Limitations of the study

In this study, in vivo extracellular recording was conducted in anesthetized mice, and anesthesia will have influenced neuronal excitability to a degree. In our recordings, ACC neurons did not respond to the mechanical stimuli induced by von Frey filament, even at high pressure (1 or 2 g). Recordings in awake, freely moving animal would be important for definitive assessments of the effect of SD on ACC neuronal activity, and the use of the same stimuli in behavioral and electrophysiological analysis would comprehensively bridge the findings. Additionally, we measured the mouse movement using activity tracking in 2 min analysis window to infer the sleep patterns of the mice before and during SD, but the analysis window was not identical to typical sleep studies using EEG/EMG recording, and our analysis could, therefore, have underestimated the inactive duration. Nevertheless, our activity tracking data clearly suggest that the RSD and FSD altered sleep patterns differentially. However, the underlying mechanisms of the differences are still unknown. Further studies are needed to reveal the mechanisms and future EEG/EMG recordings would provide more physiologically relevant sleep parameters, including actual sleep time, sleep depth (SWA intensity) and sleep pressure. Furthermore, the mechanisms of prolonged mechanical hypersensitivity and the effects of sleep irregularity on metabolism, circadian rhythm, and memory processes remain to be defined in future studies. Finally, we have not checked the effects of therapeutic agents other than mirogabalin on both pain hypersensitivity and enhanced ACC neuronal activity in RSD-hyperalgesic mice, and these will be important subjects to investigate in future studies.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Hidemasa Furue (hi-furue@hyo-med.ac.jp).

Materials availability

There is no unique reagent in this study.

Data and code availability

All data supporting the study are present in the paper or the datasets (Data S1). This article does not report the original code. Any additional information would be available from lead contact upon request.

Acknowledgments

We thank the joint-use research facilities and Center for Comparative Medicine of Hyogo College of Medicine for use of facilities. We would like to thank Editage (www.editage.jp) for English language editing. This study was supported by the following grants: JSPS KAKENHI grant JP20H04043, JP23K18441, and JP24K02806 (H.F.), JSPS KAKENHI grant JP20K16133, JP22K15206, and 25K10634 (K. Koga), 25K18708 (A.Y.) and 23K08021(H.K.), Hyogo Innovative Challenge grant (H.F. and H.K.), Hyogo College of Medicine grant for Research Promotion 2021 and 2023 (K. Koga) and 2024 (M.K.), AMED under grant nos JP24gm1510013s0102 and JP25gm1510013s0102 (H.F.), Takeda Science Foundation (K. Koga), the 45th Nakatomi Science Foundation (K. Koga), the Uehara Memorial Foundation (K. Koga), and the Naito Foundation (K.Koga).

Author contributions

K.N. designed the experiments, performed almost all the experiments, and analyzed the data. K. Koga designed and performed experiments and wrote the manuscript. M.K. originally developed the chronic sleep deprivation models and assisted with the experiments. A.Y. supported data acquisition and analysis. K. Kobayashi provided critical materials and advice regarding data interpretation. A.P. provided critical comments on the data interpretation and co-wrote the manuscript. H.K. and H.F. conceived the project, supervised the overall project, designed the experiments, and finalized the manuscript.

Declaration of interests

H.F. received funding for a collaboration study from Daiichi Sankyo Co. Ltd. The funder was not involved in the study design, collection, analysis, interpretation of data, writing of this article, or the decision to submit it for publication.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit monoclonal anti-c-Fos (9F6) antibody	Cell signaling	Cat #2250; RRID: AB_2247211
Donkey anti rabbit Cy3 antibody conjugated with Cy3	Jackson ImmunoReserch	Cat 711-165-152; RRID: AB_2307443
Bacterial and virus strains
AAVDJ-CaMKIIα-hM4D(Gi)-mCherry	This paper	N/A
AAVDJ-CaMKIIα-mCherry	This paper	N/A
Chemicals, peptides, and recombinant proteins
Neurotrace™ 640/660	Thermo Fisher Scientific	Cat #N21483
Clozapine N-oxide	Enzo Life Sciences,Inc.	Cat #BML-NS105-0025
Mirogabalin	Daiichi Sankyo	DS-5565
Diclofenac	MedChemExpress	Cat # HY-1503
Muscimol	Abcam	Cat # ab120094
Critical commercial assays
NanoTag	KISSEI COMTEC	Cat # VSMN210
Experimental models: Organisms/strains
Mouse: C57BL/6JJmsSlc	Japan SLC	N/A
Recombinant DNA
pAAV-CaMKIIα-hM4D(Gi)-mCherry	Unpublished	Cat #50477, Addgene
pAAV-CaMKIIα-mCherry	Unpublished	Cat #114469, Addgene
Software and algorithms
ClampeX version 8.2	Molecular Devices	https://support.moleculardevices.com/s/article/Axon-pCLAMP-8-Electrophysiology-Data-Acquisition-Analysis-Software-Download-Page
Clampfit version10.4	Molecular Devices	https://support.moleculardevices.com/s/article/Axon-pCLAMP-10-Electrophysiology-Data-Acquisition-Analysis-Software-Download-Page
Fiji	NIH	https://fiji.sc
Offline sorter version 3	Plexon	https://plexon.com/products/offline-sorter/
Nanotag/Viewer	KISSEI COMTEC	Cat # nanotag/Viewer
ANY-Maze	Stoelting	https://www.any-maze.com/

Experimental model and study participant details

Male C57BL/6J mice (Japan SLC) were used. All mice used were 8–12 weeks old at the start of each experiment and were housed at 22 ± 1°C with a 12-h light–dark cycle with food and water ad libitum. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of Hyogo Medical University (22-038 and 25-059AGP), performed in accordance with the institutional guidelines for animal experiments, and consistent with the ethical guidelines of the International Association for the Study of Pain.

Method details

Chronic SD models

A sleep fragmentation chamber (model 80391, Lafayette Instrument) was used to induce SD in mice.⁵⁰ The mice were acclimated to the cage and device for one week before SD without sweeper bar movement. In the FSD model, to mimic severe SD, mice are kept awake for the first 9 h of the light period and for the dark period by sweeper bar movements (every 2 min) from end to end across the bottom of the cage for 28 consecutive days. In the RSD model, to induce irregular SD observed in non-scheduled shift workers or in patients with severe sleep disorders, mice were kept awake with the same method used for the FSD model, but the time allowed to sleep (3 h) was randomly assigned to one of the following light periods (7:00–10:00, 10:00–13:00, 13:00–16:00, or 16:00–19:00 in the light period, varied from day to day) for 28 consecutive days.

Von frey test

We used the methods described in our previous study. To assess mechanical hypersensitivity, mice were placed individually in an opaque acrylic box (6 × 6 × 6 cm) on a wire mesh and habituated for approximately 1 h to allow acclimatization to the experimental environment. Calibrated von Frey filaments (0.02–2.0 g, North Coast Medical) were then applied to the plantar surfaces of the hindpaws of mice from below the mesh floor, and the 50% paw withdrawal threshold was determined using the up–down method.⁵¹ The von Frey test was performed before SD and every week for nine weeks after the start of SD. We conducted these tests between Zeitgeber time (ZT) 6-9 to mitigate the effects of circadian rhythm.⁵² To investigate the effect of ACC neuronal activity inhibition, the von Frey test was performed before and 30, 60, 90, 120, and 180 min after CNO (3 mg/kg in saline, BML-NS105-0025, Enzo Life Sciences,Inc) i.p injection. To investigate the effects of mirogabalin (3 mg/kg in saline, DS-5565, mirogabalin besylate, Daiichi Sankyo), the von Frey test was performed before and 1, 2, 3, 4, 5, 6, and 18 h after drug administration. To investigate the effects of diclofenac (10 mg/kg in saline), the von Frey test was performed before and 1, 2, 3, 4, 5, 6, and 18 h after drug administration. We conducted von Frey tests for drugs administrations between ZT 3-12.

Hot-plate test

Mice were placed on a metal surface maintained at 50°C in a clear acrylic cylinder (15 cm diameter, 30 cm height) using hot plate equipment (ND1, AS ONE). Latency to nocifensive behaviors was measured as a nocifensive end point (cut-off time: 60 s), in accordance with a previous study.⁵³

Open-field test

An open-topped white box (48 cm square) was used. At the start of the test, each animal was placed in the center of the box. The movement of the center of the mouse body was tracked for 5 min. The distance traveled, the time spent in the center zone (24 × 24 cm central area of the box), and the number of times the center zone was entered were calculated using ANY-Maze software (Stoelting).

Elevated-plus maze test

An apparatus composed of two closed arms (30 cm each), two open arms (30 cm each), and a central area (7 × 7 cm) elevated 40 cm above the floor was used. At the start of the test, mice were placed in the central area. The movement of the mouse was tracked for 5 min. The time spent in the open arms, the number of times the open arm was entered, and the distance traveled were measured using ANY-Maze software (Stoelting).

Measurement of voluntary movement

Following a previous paper,²⁷ we recorded voluntary movements with minor modifications. Briefly, under isoflurane anesthesia, 7–8 weeks old male C57BL/6N mice were implanted with a small wireless sensor (nano tag, Kissei Comtec) under the lower back skin ∼1week before the start of measurement. During the measurement, we detected voluntary movement every minute and analyzed with a 2 min window corresponding to the bar movement schedule (once every 2 min) for two days in each time point. We defined inactive periods if voluntary movements were absent during the recording periods, and defined other periods as active periods. The inactive periods in each SD cycle were calculated as sleep-like inactivity. In both the FSD and RSD model analysis, we used the values corresponding to SD cycle of the FSD protocols in the pre- phase, because the sleep periods without bar movement were randomly assigned in RSD model.

Cannula implantation and drug treatment

Following a previous paper,⁵⁴ we implanted a drug-infusion cannula above the ACC with minor modifications. Mice were deeply anesthetized with medetomidine hydrochloride (0.3 mg/kg, Domitol, Meiji Seika Pharma), midazolam (4 mg/kg, Dormicum, Astellas Pharma), and butorphanol (5 mg/kg, Vetorphale, Meiji Seika Pharma), and bilateral guide cannulae (62003, RWD o.d., 0.48 mm; i.d., 0.34 mm) were implanted above the ACC [rostrocaudal: +1.0 mm, mediolateral: 0.3 mm from bregma, dorsoventral: 0.8 mm, from the surface of the brain]. After surgery, the holes of the cannula were capped with a dummy cannula (62102, RWD), and the mice were housed individually and allowed to recover for more than 1 week. For intra-ACC infusion, a bilateral injection cannula (62203, o.d., 0.30 mm; i.d., 0.14 mm with 0.5 mm protrusion) was inserted into the guide cannula. Mirogabalin (20 μM, Daiichi Sankyo), muscimol (1 mM, ab120094, Abcam) or saline was administered bilaterally in a volume of 0.2 μL/side at a rate of 0.2 μL/min. After injection, the cannula was kept in place for 1 min to prevent backflow, and the holes of the guide cannula were recapped with a dummy cannula. Von Frey test was performed before and at 30 and 60 min after the injections.

In vivo extracellular recordings of neuronal activity in the ACC

In vivo multi-unit recordings were made from urethane-anesthetized (1.2–1.5 g/kg i.p.) mice, as described previously.^55,56 A tungsten electrode (1 MΩ, A-M System) was stereotaxically placed into the ACC through a skull burr hole using a micromanipulator (SMM-100; Narishige, Tokyo, Japan). Multi-unit neuronal firing was amplified, bandpass-filtered at 300–3000 Hz, and stored on a personal computer at 20 kHz using pCLAMP 10 (Molecular Devices). Pinch stimulation was manually applied 3 times to the contralateral hindpaw with toothed forceps for 10 s. Frequencies of baseline (for 10 s before the stimulation) and pinch response (during the stimulation), and after discharge (for 10 s and 10-20 s after the end of stimulation) in three trials were averaged.^57,58 The recorded multiunit signals were spike-sorted (offline sorter version 3; Plexon), and spike width was calculated by the trough to peak interval of the mean spike waveform. We defined units with spike widths of >0.45 ms and <0.35 ms as putative pyramidal neurons and interneurons, respectively⁵⁹ and used only the data from putative pyramidal neurons for analysis. CNO (3 mg/mL) was administered i.p. during recording. Mirogabalin (3 mg/kg) was i.p. administered approximately 4 h before recording.

Adeno-associated virus (AAV) production and purification

The plasmid pAAV-CaMKIIα-hM4D(Gi)-mCherry (Addgene, #50477)²⁹ and pAAV-CaMKIIα-mCherry (Addgene, #114469) were kindly gifted from Bryan Roth lab and Karl Deisseroth lab, respectively. AAV_DJ-CaMKIIα-hM4D(Gi)-mCherry and AAV_DJ-CaMKIIα-mCherry was produced using the AAV Helper-Free System (Agilent Technologies) as previously described.⁶⁰

Microinjections

We used a previously reported method with some modifications.⁶¹ Mice were deeply anesthetized with medetomidine hydrochloride (0.3 mg/kg, Domitol, Meiji Seika Pharma), midazolam (4 mg/kg, Dormicum, Astellas Pharma), and butorphanol (5 mg/kg, Vetorphale, Meiji Seika Pharma), and their heads were fixed in a stereotaxic apparatus (SR-5M-HT, Narishige). rAAV (AAV_DJ-CaMKIIα-hM4D(Gi)-mCherry or AAV_DJ-CaMKIIα-mCherry, 2.0 × 10¹² GC/mL diluted with PBS) were bilaterally injected (approximately 250 nL in one site) into the ACC [rostrocaudal: +1.0 mm, mediolateral: 0.3 mm from bregma, dorsoventral: 0.8 mm, from the surface of the brain]. We performed microinjections one week before the start of acclimation to the sleep fragmentation chamber.

Slice electrophysiology

Slice electrophysiology experiments were performed as previously described.⁵⁵ Mice were deeply anesthetized with medetomidine hydrochloride (0.3 mg/kg), midazolam (4 mg/kg), and butorphanol (5 mg/kg), and the brain was quickly removed and placed into cold high-sucrose artificial cerebrospinal fluid (aCSF) (250 mM sucrose, 2.5 mM KCl, 2 mM CaCl₂, 2 mM MgCl₂, 1.2 mM NaH₂PO₄, 25 mM NaHCO₃, and 11 mM glucose). Coronal brain slices (300-μm thick) were cut using a vibrating microtome (NLS-MT, Dosaka), and then the slices kept in oxygenated aCSF solution (125 mM NaCl, 2.5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 1.25 mM NaH₂PO₄, 26 mM NaHCO₃ and 20 mM glucose) at room temperature (22–25°C) for at least 30 min.

Individual slices were then placed in a recording chamber that was continuously perfused with an aCSF solution at room temperature. Patch-clamp recordings were made from single neurons and visualized using infrared differential interference contrast optics (BX50WI; Olympus). The patch pipettes (4–7 MΩ) were filled with an internal solution (K-gluconate 135, CaCl₂ 0.5, MgCl₂ 2, KCl 5, EGTA 5, 5 Mg-ATP, and HEPES, 0.2% Neurobiotin, pH 7.2 adjusted with KOH). Signals were amplified with MultiClamp 700A amplifier and pCLAMP 10.4 acquisition software (Molecular Devices, USA), and digitized with an analog-to-digital converter (Digidata 1321A, Molecular Devices), stored on a computer using a data acquisition program (Clampex version 8.2, Molecular Devices) and analyzed using a software package (Clampfit version10.4, Molecular Devices). Membrane potentials were recorded in the current-clamp mode, and the discharge patterns of the recorded neurons were examined by passing depolarizing current pulses through the recording electrode from the resting membrane potential.

Immunohistochemistry and confocal imaging

Immunohistochemical experiments were performed according to the methods described in our previous study.⁶¹ Mice were deeply anesthetized with urethane (1.2–1.5 mg/kg) and perfused transcardially with phosphate buffered saline (PBS), followed by ice-cold 4% paraformaldehyde/PBS. The brains were removed, postfixed in the same fixative for overnight at 4°C, and placed in 30% sucrose solution for two overnight at 4°C. Coronal brain sections (50-μm thick) were made by cryostat (CM3050S, Leica). For c-Fos staining, a rabbit monoclonal anti-c-Fos antibody (1: 1000, 9F6, #2250, Cell Signaling) and a donkey anti-rabbit Cy3 (1:500, AB_2307443, Jackson ImmunoReserch) were used. For Nissl staining, Neurotrace 640 (N21483, Thermo Fisher Scientific) was used at 1:100 dilution following the manufacturer’s protocol. Immunofluorescent images were obtained using a confocal laser microscope (LSM780, Carl Zeiss or LSM900, Carl Zeiss). C-Fos-positive cells were quantified using ImageJ Fiji software (https://fiji.sc/) and averaged 3 of slices from each mouse were aggregated for each group.

Quantification and statistical analysis

Statistical analyses were performed using Prism 9 (GraphPad). All data are shown as the mean ± SEM. Statistical significance was determined by one-way ANOVA with Dunnett's multiple comparison test, one-way ANOVA with Tukey's multiple comparison test, two-way repeated-measures ANOVA with Bonferroni's multiple comparison test, unpaired t-test, and unpaired t-test with Welch’s correction. Differences were considered statistically significant at p< 0.05.

Published: January 20, 2026