Downregulation of ciRNA‐ Kat6b in dorsal spinal horn is required for neuropathic pain by regulating Kcnk1 in miRNA‐26a‐dependent manner

Abstract Aims Nerve injury‐induced maladaptive changes in gene expression in the spinal neurons are essential for neuropathic pain genesis. Circular RNAs (ciRNA) are emerging as key regulators of gene expression. Here, we identified a nervous‐system‐tissues‐specific ciRNA‐Kat6 with conservation in humans and mice. We aimed to investigate whether and how spinal dorsal horn ciRNA‐Kat6b participates in neuropathic pain. Methods Unilateral sciatic nerve chronic constrictive injury (CCI) surgery was used to prepare the neuropathic pain model. The differentially expressed ciRNAs were obtained by RNA‐Sequencing. The identification of nervous‐system‐tissues specificity of ciRNA‐Kat6b and the measurement of ciRNA‐Kat6b and microRNA‐26a (miRNA‐26a) expression level were carried out by quantitative RT‐PCR. The ciRNA‐Kat6b that targets miRNA‐26a and miRNA‐26a that targets Kcnk1 were predicted by bioinformatics analysis and verified by in vitro luciferase reports test and in vivo experiments including Western‐blot, immunofluorescence, and RNA–RNA immunoprecipitation. The correlation between neuropathic pain and ciRNA‐Kat6b, miRNA‐26a, or Kcnk1 was examined by the hypersensitivity response to heat and mechanical stimulus. Results Peripheral nerve injury downregulated ciRNA‐Kat6b in the dorsal spinal horn of male mice. Rescuing this downregulation blocked nerve injury‐induced increase of miRNA‐26a, reversed the miRNA‐26a‐triggered decrease of potassium channel Kcnk1, a key neuropathic pain player, in the dorsal horn, and alleviates CCI‐induced pain hypersensitivities. On the contrary, mimicking this downregulation increased the miRNA‐26a level and decreased Kcnk1 in the spinal cord, resulting in neuropathic pain‐like syndrome in naïve mice. Mechanistically, the downregulation of ciRNA‐Kat6b reduced the accounts of miRNA‐26a binding to ciRNA‐Kat6b, and elevated the binding accounts of miRNA‐26a to the 3′ untranslated region of Kcnk1 mRNA and degeneration of Kcnk1 mRNA, triggering in the reduction of KCNK1 protein in the dorsal horn of neuropathic pain mice. Conclusion The ciRNA‐Kat6b/miRNA‐26a/Kcnk1 pathway in dorsal horn neurons regulates the development and maintenance of neuropathic pain, ciRNA‐Kat6b may be a potential new target for analgesic and treatment strategies.


| INTRODUC TI ON
Nerve injury-induced neuropathic pain is a chronic and refractory disease, therefore affecting the quality of life of many patients. It is estimated that neuropathic pain in the general population is to have a prevalence ranging between 3% and 17%. 1 In the United States, over 600 billion dollars per year is spent on healthcare costs related to neuropathic pain management. 2 However, as current medications such as opioids and nonsteroidal anti-inflammatory drugs are ineffective or have severe side effects in most neuropathic pain patients, the therapeutic effect is limited. 3 Neuropathic pain is characterized by abnormal hypersensitivity to stimuli (hyperalgesia) and nociceptive responses to non-noxious stimuli (allodynia). Central (e.g., spinal cord dorsal) sensitization is thought to be a vital role in pain hypersensitivity. These abnormal spinal sensitization activities are responsible for peripheral nerve injury-induced maladaptive changes in such pain-associated genes as an ion channel, receptor, and intercellular signal molecular in the spinal cord. [4][5][6][7][8][9] Exploring the regulation mechanism of dysfunctional spinal genes may provide a new avenue for neuropathic pain management. Circular RNAs (circRNAs), as a kind of novel identified noncoding RNAs, have attracted great attention due to their potent and multifunction in the regulation of gene expression. 10,11 A large number of circRNAs are dysregulated in the spinal cord following peripheral nerve injury, 12,13 but the molecular mechanism underlying neuropathic pain is still poorly understood. We identified a nervous system-specifically expressed circRNA -ciRNA-Kat6b -is highly enriched in the spinal cord and decreases the expression in the spinal cord after peripheral nerve injury. But it remains unclear the mechanism of ciRNA-Kat6b involved in neuropathic pain.
Potassium channels control the excitability of spinal neurons, therefore, become a critical player in the occurrence of pain hypersensitivity. 14 KCNK1 belongs to the two-pore domain background potassium (K2P) channel family and is the first identified member in this family. Kcnk1 is enriched in the peripheral and central nervous systems such as dorsal root ganglion, trigeminal nerve, and spinal cord, 15,16 peripheral nerve injury decreases the level of Kcnk1 in mouse DRG. 17 Rescuing this decrease alleviates nerve injuryinduced mechanical, thermal, and cold pain hypersensitivities. Kcnk1 has become a potential modulation factor in the development and maintenance of neuropathic pain. 17 Here, we demonstrated that the decreased of ciRNA-Kat6b contributes to development and maintenance of nerve injury-induced neuropathic pain by regulating microRNA-26a (miR-26a)-triggered Kcnk1 in spinal cord neurons.
ciRNA-Kat6b may be a critical player in neuropathic pain.

| Animals and pain model
Adult male Kunming mice (20-25 g) were used in this study. All animal procedures were approved by the animal care committee of Xuzhou Medical University (Xuzhou, China). All animals were kept at 25°C and 40% humidity under a 12:12 h light/dark cycle and with ad libitum access to food and water. Chronic inflammatory pain was induced by subcutaneous administration of CFA (40 μl; F5881; Sigma-Aldrich) into the plantar surface of the left hind paw. The unilateral sciatic nerve chronic constrictive injury (CCI) model was performed as described previously. 4 All mice were maintained in a warm environment until they recovered from anesthesia.

| Behavioral tests
Before the behavior testing, the locomotor function was tested according to the previous method. 9 Thermal hyperalgesia and mechanical allodynia were measured, respectively, as previously described. 8 Briefly, thermal nociceptive behavior was assessed by measuring mouse paw-withdrawal latency in response to a thermal stimulus with an analgesia meter (IITC Model336 Analgesia Meter, Series 8; IITC Life Science). The time required for the stimulus to elicit withdrawal of the hind paw was recorded. Von Frey (Stoelting) filaments that produce different forces were applied to detect mechanical allodynia. Starting with a 0.16 g and ending with a 6.0 g filament. In the absence of a paw withdrawal response, a stronger stimulus was presented; when paw withdrawal occurred, the next weaker stimulus was chosen. The optimal threshold filaments were presented five times, respectively, at 5 min intervals. All behavioral tests were performed in a double-blind trial fashion in this study. KYCX21_2707; Key project of the Natural Science Foundation of Jiangsu Education Department, Grant/Award Number: 22KJA320008 of Kcnk1 mRNA, triggering in the reduction of KCNK1 protein in the dorsal horn of neuropathic pain mice.

Conclusion:
The ciRNA-Kat6b/miRNA-26a/Kcnk1 pathway in dorsal horn neurons regulates the development and maintenance of neuropathic pain, ciRNA-Kat6b may be a potential new target for analgesic and treatment strategies.

| Spinal and DRG tissue collection
Mice were anesthetized with isoflurane and the spinal cord within the lumbar segments (L3-L5) was removed rapidly. The DRG of the spinal lumbar segment (L3/4 DRGs) was extracted. All operations were performed on the ice box and put into snap-frozen in liquid nitrogen, and stored at −80°C. RNase R treatment, 5 μg of RNA was incubated with 4 U/μg of RNase R (Epicenter) for 20 min at 37°C and purified by phenol-chloroform extraction. GAPDH and U6 as internal control. The expression levels of the target genes were quantified relative to Gapdh or U6 expression (cycle threshold [Ct]) using the 2 −∆∆CT methods.

| Spinal neuron cell culture
The culture of spinal neuron cells was carried out described previously. 8 Briefly, 3 to 4 days-old mice under deep anesthesia were decapitated and the lumbar L3 to L5 segments of the dorsal spinal cord were collected. After enzymatical digestion with papain and mechanical dissociation, the homogenate was centrifugated for 5 min at 500 rpm. After centrifugated, the supernatant was removed and replaced with 5 mL of culture medium, the composition of which was as follows: MEM-(Invitrogen), FCS (5% v/v; Invitrogen), heat-inactivated horse serum (5% v/v; Invitrogen), penicillin, and streptomycin (50 IU/ mL for each; Invitrogen), transferrin (10 mg/mL; Sigma-Aldrich), insulin (5 mg/mL; Sigma-Aldrich), putrescine (100 nM; Sigma-Aldrich), and progesterone (20 nM; Sigma-Aldrich). After trituration with a fire-polished Pasteur pipette, the cells were plated on plastic culture dishes. Cultures were maintained in a water-saturated atmosphere (95% air, 5% CO 2 ) at 37°C until used (10-15 days). Two days after the cells were seeded, cytosine arabinoside was added to the culture medium for 24 h to reduce glial proliferation.

| Spinal astrocytes and microglia cultures
The isolation of spinal astrocytes and microglia cells was performed as described previously with few modifications. 8

| SiRNA, mimics and inhibitor
The siRNA, mimics, and inhibitor were used for intrathecal injection as described previously. 5 In brief, holding the mouse by the

| Immunofluorescence and fluorescence in suit hybridization (FISH)
To

| Statistical analysis
All data were first tested for normality using a Shapiro-Wilk test normality test by Prism GraphPad 8.0. The data were presented as Means ± SEM. One-way or two-way ANOVA or paired or unpaired Student's t-test were used to statistically analyzed. When ANOVA showed a significant difference, pairwise comparisons between means were tested by the post hoc Tukey method. Statistical analyses were performed with Prism (GraphPad 8.0). p < 0.05 was considered statistically significant in all analyses.

| Identification of spinal ciRNA-Kat6b and its specificity in the nervous system
To explore the circRNA mechanisms underlying neuropathic pain, we carried out the deep sequencing of circRNAs from the ipsilateral spinal dorsal horn of mice subjected to chronic constriction injury (CCI) and Sham surgery. Among the differential expression ciRNAs (Table S1), ciRNA-Kat6b was downregulated by 2.1 folds after peripheral nerve injury, this downregulation was further confirmed by qRT-PCR ( Figure 1A). Interestingly, although ciRNA-Kat6b was not the most differential expression ciRNA, ciRNA-Kat6b was specifically expressed in nervous system tissues including 9 mouse nervous tissues (including the spinal cord, dorsal ganglia root, thalamus, cerebellum, hippocampus, trigeminal ganglion, cortex, and brain stem), but undetected in 5 nonnervous tissues (including heart, lung, liver, kidney, and spleen) The results showed that ciRNA-Kat6b in neurons was 5.9-fold as that in microglial cells and 4.3-fold as that in astrocyte cells.
These results together with FISH data supports the conclusion that ciRNA-Kat6b was mainly expressed in spinal cord neurons ( Figure 1F). Because the cultured neurons can be depolarized with high-concentration KCl to mimic sensitized in vivo neurons by nociceptive response, 19 we treated the cultured spinal neurons with 50 mM KCl for 12 h and found that ciRNA-Kat6b was significantly Further analysis by the cirbase database showed that ciRNA-Kat6b was conserved in mammal animals such as humans, rats, and dogs ( Figure S1A). CiRNA-Kat6b was also detected in the human spinal cord and DRG ( Figure 1H). Collectively, we identified that ciRNA-Kat6b is specifically expressed in the nervous system, and has high conservation between mice and humans.

| Mimicking CCI-induced spinal ciRNA-Kat6b decrease leads to nociceptive hypersensitivity
We injected with LV-ciR-Kat6b-shRNA, but not LV-Scr, displayed both evoked thermal and mechanical sensitivities on 5, 7, and 9 after injection ( Figure 3J,K). As expected, no changes in locomotor function (Table 1) were found in either siRNA or shRNA-injected mice. Spinal ciRNA-Kat6b downregulation likely caused neuropathic pain-like symptoms.

| CiRNA-Kat6b acts to be a sponge absorbing miRNA-26a
How to ciRNA-Kat6b in luciferase promoter, significantly decreased the activity of the pGL6 reporter by 33% compared with the mutated pGL6 reporter ( Figure 4C). Contrarily, the co-transfection of miRNA-26a inhibitor (Ih), a synthesized small RNA with its reverse complementary sequence used to knockdown miRNA-26a, increased the activity of the wild reporter by 55%, compared with the mutated reporter ( Figure 4C). Furthermore, we wondered whether neuropathic pain stimulus could regulate miRNA-26a expression. We found that nerve injury upregulated the expression of spinal miRNA-26a from day 3 to 21 after CCI surgery ( Figure 4D). These data indicate that ciRNA-Kat6b likely acts to be a sponge absorbing miRNA-26a in spinal neurons.

| CiRNA-Kat6b regulates neuropathic pain by targeting miRNA-26a
To explore whether miRNA-26a could mediate the neuropathic pain intrathecally injected into the mice subjected to 7-day CCI surgery.
Scrambled small RNA (Scr) was used as a control. The mice injected with miRNA-26a-Ih, but not Scr, showed the attenuated response in paw withdrawal latencies to thermal and mechanical stimuli on day 1 after injection ( Figure 5A,B). Similarly, the injected mice with LV-miRNA-26a shRNA (miR-26a shRNA), producing miRNA-26a Two consecutive days of intrathecal injection of Lenti-miRNA-26a (Lenti-miR-26a) or control scrambled virus (Lenti-Scr) into the spinal cord increased the hypersensitivity to heat stimuli (I) and to von Frey filaments stimuli (J) at the different days after lentivirus injection. n = 8 mice/ group. *p < 0.05, **p < 0.01 versus the Lenti-Scr-treated mice at the corresponding time points by two-way ANOVA with repeated measures followed by post hoc Tukey test. The red arrow indicates Lenti-miR-26a or Lenti-Scr. (K, L) Two consecutive days intrathecal injection of miRNA-26a inhibitor (26a-Ih) on day 3 after intrathecal injection of Lenti-ciRNA-Kat6b-shRNA (LV-ci-K6b-shR) or its scrambled virus control (LV-Scr) in naïve mice inhibited the production of pain hypersensitivity to heat stimuli (K) and to von Frey filaments stimuli (L). n = 8 mice/ group. *p < 0.05, **p < 0.01 versus the LV-Scr plus the scrambled inhibitor (Scr) group; # p < 0.05 versus the LV-ci-K6b-shR plus Scr group by two-way ANOVA with repeated measures followed by post hoc Tukey test. The red arrow indicates LV-ci-K6b-shR or LV-Scr injection. Blue arrows indicate 26a-Ih or its control.
inhibitor, but not its control LV-Scr, displayed the alleviated pain hypersensitivity to thermal and mechanical stimulus ( Figure 5C,D) from days 2 to 7 after 2 consecutive days of intrathecal injection.
However, locomotor impairment, measured by reflex tests including placing, grasping, and righting as previously described 9 was not observed after synthesized or lentivirus-producing miRNA-26a inhibitor ( Table 1

| Upregulation of miRNA-26a decreases the expression of Kcnk1 in spinal cord
To explore how spinal miRNA-26a increase is involved in neuropathic pain, we predicted its downstream targets by combining seven independent programs ( Figure 6A). As the potassium channels paly a critical role in the regulation of neuronal excitability under chronic pain conditions, 13,20 we focused on the potassium channels among the predicted targets by miRNA-26a. We found 9 potassium channel genes might become miRNA-26a's potential downstream targets ( Figure 6A). Among them, Kcnk1 (one member of the two-pore domain background potassium (K2P) channel family) is likely the target of miRNA-26a due to its highest frequency appearance by 5 prediction programs ( Figure 6A).
The 3′UTR of mouse Kcnk1 mRNA constitutes a conserved miR-26a binding sequence of "UACUUGAA" (220-227 bp, the first base of 3′UTRas +1), and this region displayed a high specific conversation among 4 mammal specifics including mouse, rat, human, and chimp ( Figure 6B). KCNK1 is a critical player in the modulation of the neuropathic pain through controlling K + current in peripheral sensory neurons. 17 Therefore, Kcnk1 was chosen as a potential target of miRNA-26a in this study. Our findings showed that peripheral nerve injury also decreased Kcnk1 mRNA expression in the ipsilateral dorsal spinal cord from the day 3 to day 21 after surgery ( Figure 6C). The KCNK1 protein in the ipsilateral spinal dorsal cord was also decreased on day 7 after CCI surgery ( Figure 6D). Additionally, single-cell PCR showed that Kcnk1, miRNA-26a, and ciRNA-Kat6b were co-expressed in the spinal neuron ( Figure 6E). Thus, we wondered whether miRNA-26a could regulate the expression of KCNK1 in the dorsal spinal cord. Further studies between the in vitro and in vivo experiments were carried out to verify this point. In vitro, we cloned the 3′UTR of mouse Kcnk1 mRNA containing the region bound by miRNA-26a into the firefly luciferase reporter (CHK-wt-kcnk1).
As expected, the luciferase assay revealed that co-transfection of wild-type reporter (CHK-wt-kcnk1), but not control mutated reporter (CHK-mut-kcnk1), with miRNA-26a mimics significantly decreased the activity of the luciferase in HEK293T cells ( Figure 6F). Contrarily, the co-transfection of wild-type but not mutated reporter with miRNA-26a inhibitor increased the activity of the luciferase ( Figure 6F)

| The decrease of Kcnk1 is required for the pain hypersensitivity induced by miRNA-26a
To further investigate whether spinal Kcnk1 is involved in the development of neuropathic pain, we used the lentivirus to overexpress indicating the pain-like behavior genesis. Expectedly, the locomotor injury was not affected by siRNA or scramble control (Table 1).
Finally, to test whether miRNA-26a modulates pain hypersensitivity via the mediation of Kcnk1, we increased the expression of Kcnk1 post or before miRNA-26a in the spinal cord of naïve mice.
We found that the intrathecal pre-injection of Lenti-Kcnk1 on day 3 before miRNA-26a mimics injection prevented the miRNA-26a increase-induced nociception sensitivity to thermal and mechanical stimulus in naïve mice ( Figure 7G,H). Similarly, the intrathe- abolished the miRNA-26a upregulation-induced pain sensitivity in naïve mice ( Figure 7I,J). These data suggested that Kcnk1 mediates the pain hypersensitivity induced by the miRNA-26a increase.

| DISCUSS ION
circRNAs have been emerging to be a novel regulatory function mechanism in gene expression and attract widespread attention for their key roles in myriad biological processes and human diseases. 21,22 Bioinformatic analyses from human brain tissues identified a large number of neuronal-specific circRNAs, 80% of mouse brain circRNAs are also detected in the human brain. The expression levels of these circRNAs are often changed during neuronal differentiation and in nervous system-related diseases, 23  primary nerve tissue such as DRG 25 to the central nervous system such as spinal cord 26 and pain-related brain regions. 27 MiRNA-26a is involved in several diseases as neonatal sepsis, 28 stroke, 29 coronary heart disease (CHD), 30 and cancer. 31