Male Sprague-Dawley (SD) rats (∼200 g) were obtained from the Experimental Animal Central of Central South University. The animal experiments were approved by the Animal Care and Use Committee of Central South University. Spinal nerve ligation (SNL)-induced neuropathic pain model was established as previous described (Zhang et al., 2013). Briefly, following anesthetized with 1% sodium pentobarbital (40 mg/kg i.p.), the fifth lumbar (L5) spinal nerves were isolated and ligated tightly with 6–0 silk suture. For sham-operated mice, rats received identical surgical procedure without ligation. Mechanical withdrawal threshold (MWT) and thermal withdrawal latency (TWL) were used to validate the SNL model.

To test the effect of lentivirus shPKIA-AS1 (Lv-shPKIA-AS1, sequence: CCGGGACGACCCTGACCATGGTCGCGATATCTCGAGATATTTGAGGTCAGGGTCGTCAATTTG, Sigma Aldrich), SNL rats were received intrathecal injections of either Lv-shPKIA-AS1 (MOI = 100) or the lentivirus with scramble sequence on the third day after surgery.

To assess the effect of PKIA-AS1 overexpression, rats were received intrathecal injections of either Lv-PKIA-AS1 [multiplicity of infection (MOI) = 100] or the lentivirus with scramble sequence on the fifth day of observation.

To assess the effect of CDK6 overexpression, rats were received intrathecal injections of either Lv-CDK6 (MOI = 100) or the lentivirus with scramble sequence on the third day after surgery. Intrathecal injection process was performed as previously described (Zhang et al., 2013). Briefly, rats were placed in a prone position after anesthesia. After creating a small opening at the intervertebral space between the L4–L6 vertebrae, a sterile PE10 intrathecal catheter was inserted through the opening into the lumbar enlargement. After catheterization, all rats were allowed to recover for two days before being used in other experiments.

Mechanical allodynia was evaluated by paw withdrawal threshold (PWT) using von Frey filaments. Briefly, rats were put into a transparent plastic box (22 cm × 12 cm × 22 cm). The box has a metal mesh floor. The calibrated von Frey filaments (IITC, Woodland Hills, CA, United States) were used to make pressure on the plantar surface of rat hind paw. The investigators recorded the size of the filaments when paw withdrawal. Paw withdrawal latency (PWL) was used to evaluated thermal hyperalgesia by the Plantar Test Instrument (Hargreave’s Method). The hind paws were tested alternately at 5-min intervals. The investigators recorded the duration between stimuli and paw withdrawal. The cut-off time was at 30 s. All tests were performed blindly.

On the 15th day after spinal nerve ligation (SNL) or sham operation, rats were anesthetized with pentobarbital sodium (150 mg/kg IV) and their spinal cord tissues (L5) were removed and frozen by liquid nitrogen immediately. RNeasy Mini Kit (Qiagen, GmBH, Hilden, Germany) was used to extract total RNA from spinal cord tissues according to the manufacturer’s instructions. NanoDrop 2000 spectrophotometer was used to quantify purified total RNA. To analyze lncRNA expression profiles (Arraystar Rat LncRNA Expression Array V2.0), the purified total RNA was sent to Novogene Co. Ltd. (Beijing, China). Differentially expressed lncRNAs were identified through fold change (>2) and P-value (<0.05).

TRIzol reagent (Invitrogen) was used to extract total RNA from cells or tissues. Maxima First Strand cDNA Synthesis kit (cat no. K1642; Thermo Fisher Scientific, Inc.) was used for reverse transcription according to the manufacturer’s protocol. Quantitative PCR amplification was performed with an CFX96 Touch^TM Deep Well Real-Time PCR Detection System (Bio-Rad, Hercules, CA, United States). Expression of lncRNAs, PKIA-AS1, IL-1β, IL-6, IL-12, TNFα, and CDK-6 was detected via UltraSYBR Mixture (cat. no. CW2602; CWBio) according to the manufacturer’s protocol. QPCR was performed at the condition: 95.0°C for 3 min, and 39 circles of 95.0°C for 10 s and 60°C for 30 s. Expression of β-actin was used as an endogenous control. Relative gene expression was calculated using 2^-ΔΔCt method. The real-time PCR primer sequences are shown in Supplementary Table S1.

Radio immunoprecipitation assay (RIPA) lysis buffer (Boster, Wuhan, China) was used to extract protein from spinal cord tissues (L5). BCA Protein assay kit (Thermo Scientific, Waltham, MA, United States) was used to measure protein concentrations. Proteins (60 μg) were separated by 10% SDS/PAG. The membranes were blocked by 5% non-fat milk for 30 min at room temperature and then immunoblotted with the following primary antibodies: CDK6 antibody (cat no. 13331, Rabbit monoclonal, diluted at 1:1000), DNMT1 antibody (cat no. 5032, Rabbit monoclonal, diluted at 1:1000), DNMT3A antibody (cat no. 32578, Rabbit monoclonal, diluted at 1:1000), DNMT3B antibody (cat no. 57868, Rabbit monoclonal, diluted at 1:1000), GAPDH antibody (cat no. 5174, Rabbit monoclonal, diluted at 1:1000). All antibodies were purchased from Cell Signaling Technology. Membranes were then incubated with peroxidase-conjugated secondary antibody, and specific bands were detected with a Bio-Rad (Hercules, CA, United States) imaging system.

Spinal cord tissues were cut into twenty-μm thick sections for immunofluorescence assays. After permeabilized with 0.5% Triton-100, the sections were blocked with BSA for 60 min at 37°C. The sections then were incubated with primary antibody (cat no. ab68428, anti-GFAP, 1:500; cat no. ab15690, anti-Iba-1, 1:200, Abcam, Cambridge, United Kingdom) at 4°C overnight. After that, sections were washed by PBS and secondary antibody was added to incubation at 37°C for 1 h. The coverslips were stained with DAPI (1:1000, Santa Cruz Biotechnology) for 2 min at room temperature and mounted. Fluorescence microscope (Nikon ECLIPSE 80i, Nikon Corporation, Tokyo, Japan) was used to immunofluorescence acquire images. ImageJ (version 1.8) was used to quantify the fluorescence intensity on 8 sections per animal at 40× magnification.

Total proteins were extracted and purified from fresh spinal cord tissues (L5) of Lv-PKIA-AS1 infected mice and the control mice using a ReadyPrep^TM Protein Extraction kit (Bioscience) following the procedure recommended by the manufacturer. The proteomics analysis was performed as previously described (Yang et al., 2014). Protein samples were separated by two-dimensional polyacrylamide gel electrophoresis (2-DE). Samples containing 150 μg of protein were diluted to 450 μl with rehydration solution and used for isoelectric focusing. The proteins were electrophoresed in SDS–PAGE and then stained with coomassie brilliant blue dye. Spot-detect and determine the quantity were analyzed by DeCyder software version 6.5 (GE). The statistical significance was assessed by using a one-way ANOVA analysis. Protein spots were selected as the mean ratio was greater than 1.5-fold or less than -1.5-fold.

PC12 cells were purchased from the American Type Culture Collection (Manassas, VA, United States). The cells were cultured in DMEM medium (Gibco) supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA, United States) and 1% penicillin/streptomycin in a 5% CO2 incubator at 37°C.

The cells were infected with the lentiviral transduction particles for PKIA-AS1. Briefly, when the cells confluence reached 80–90% in 6-well plates, the cells were infected with lentivirus [multiplicity of infection (MOI) = 50] in the presence of 5 mg/ml polybrene (Sigma-Aldrich, St. Louis, MO, United States) for 48 h.

To test the promoter activity, CDK6 promoter was cloned into the pGL3-Basic vector using the Fast-Fusion^TM Cloning Kit (FulenGen, Guangzhou, China). Luciferase reporter constructs were co-transfected with pRL-TK (Promega) into PC12 cells, followed by infection with Lv-PKIA-AS1. Dual Luciferase Reporter Assay Kit (Promega) was used to measure the luciferase activity according to the manufacturer’s instruction.

RNA-binding protein immunoprecipitation assay was conducted using Magna RIP Kit (EMD Millipore, Billerica, MA, United States) according to the manufacturer’s instruction. After the cells were lysed by RIP lysis buffer, magnetic beads conjugated to human anti-Ago2 antibody (Millipore) or control antibody (normal mouse IgG; Millipore) were added into cell lysate for incubation at 4°C overnight. CDK6 and PKIA-AS1 expression were measured by qRT-PCR.

CDK6 promoter (-2000–+150) methylation status was measured by MSP. Genomic DNA (1 μg per sample) extracted by the Qiagen FFPE DNA Kit (Qiagen, CA, United States) was modified with bisulfite using the EZ DNA Methylation-Gold Kit (Zymo, Orange County, CA, United States) according to the manufacturer’s instructions. Bisulfate-treated DNA was used to perform quantitative methylation-specific PCR (MSP). The qPCR steps were as following: initial denaturation at 95.0°C for 3 min; 39 cycles of 95.0°C for 10 s and 60°C for 30 s. The primers were as following: methylated-specific primer, forward, 5′-AGGCGGTTGTAGTTTTTGTAGTC-3′, reverse, 5′-ATTATTATTATTACTTTCCCACGCT-3′; unmethylated-specific primer, forward, 5′-GGAGGTGGTTGTAGTTTTTGTAGTT-3′, reverse, 5′-ATTATTATTATTACTTTCCCACACT-3′.

Rat IL-1β, IL-6, IL-12, and TNFα ELISA kit was purchased from Cwbiotech (Beijing, China). The lumbar spinal cord segments (L5) were removed and were homogenized in a lysis buffer. ELISA was performed according to manufacturer’s protocol.

The experiments repeated at least three times. Data are expressed as means ± SEM. Statistical analyses were processed on GraphPad Prism software (GraphPad Software Inc., La Jolla, CA, United States). Student’s t-test and one-way ANOVA with Bonferroni test were used to assess the statistical significance of the differences between two groups and multiple groups, respectively. Two-way repeated measure ANOVA with Bonferroni post-test for multiple comparisons at each time point was used to analyze PWT and PWL. P less than 0.05 was considered statistically significant.

To investigate the role of lncRNAs in neuropathic pain induced by spinal cord injury, we established a SNL rat model (Figure 1A,B). Total RNA was extracted from spinal cord tissues of these SNL rats and sham-operated control rats to perform lncRNA array analysis. We successfully identified 1259 downregulated lncRNAs and 2473 upregulated lncRNAs in SNL rats compared to controls (Figure 1C). Table 1 presents the detail information on the top 10 most upregulated and top 10 most downregulated lncRNAs. We then investigated the profiles of the 10 most highly upregulated lncRNAs in the spinal cord of SNL rats by real-time quantitative PCR. These substantially upregulated lncRNAs, included PKIA-AS1 (fold change >15), SNHG4 (fold change >5), STAC3 (fold change >5), and CIRBP-AS1 (fold change >5) (Figure 1D). Specific overexpression in the SNL model strongly suggests that these differentially expressed lncRNAs are involve in the pathogenesis of neuropathic pain. In this study, we focused on the contributions of PKIA-AS1 to SNL-associated NP development as it exhibited the greatest upregulation.

Similar to NP development, expression of PKIA-AS1 increased over time following SNL (Figure 2A). Therefore, we investigated whether reducing PKIA-AS1 expression in SNL rats by shRNA knockdown can attenuate SNL induced neuropathic pain. Real-time PCR demonstrated that PKIA-AS1 expression in spinal cord of SNL rats was markedly downregulated 3 days following infection with a lentivirus vector targeting PKIA-AS1 (Lv-PKIA-AS1) compared with the SNL rats infected with control lentivirus (P < 0.05, N = 5; Figure 2B).

The effects of PKIA-AS1 shRNA on both mechanical and thermal hyperalgesia were tested by measuring the paw withdrawal threshold (PWT) and latency (PWL). Animals developed progressive mechanical hyperalgesia following SNL as evidenced by reduced PWT that plateaued on the 7th day after surgery. At 3 days post-surgery, PWT was significantly lower in rats earmarked for Lv-NC and Lv-shPKIA-AS1 infection (SNL+Lv-NC and SNL+Lv-shPKIA-AS1 groups, respectively) compared to sham-operated controls. The PWT of the SNL+Lv-shPKIA-AS1 group rose significantly above that in the SNL+Lv-NC group following infection (P < 0.05, N = 8 rats per group), indicating reduced hyperalgesia. However, the PWT of the SNL+ Lv-shPKIA-AS1 group remained slightly lower than that of the control group (Figure 2C). Similarly, the PWL was significantly shorter in the SNL+Lv-NC, and SNL+ Lv-shPKIA-AS1 groups compared with the control group on the 3th day after surgery, but was significantly higher in the SNL+Lv-shPKIA-AS1 group than the SNL+Lv-NC group following infection (P < 0.05, N = 8 rats per group) (Figure 2D). These results suggest that knockdown of PKIA-AS1 attenuates SNL-induced neuropathic pain.

Next, we assessed whether PKIA-AS1 overexpression alone is sufficient to cause neuropathic pain–like symptoms in uninjured control rats. Seven days after infection with Lv-PKIA-AS1, real-time PCR showed that expression of PKIA-AS1 in the spinal cord was fivefold higher than in control rats or rats infected with the control lentivirus (P < 0.05, N = 5) (Figure 3A). The PWT in rats infected with Lv-PKIA-AS1 started to decrease by the second day post-infection and was even lower on days 4, 6, and 9 (P < 0.01, N = 8 rats per group) compared with rats infected with control lentivirus (Figure 3B), indicating progressive development of hyperalgesia. Similarly, PWL was significantly shorter in rats following Lv-PKIA-AS1 infection compared to rats infected with the control lentivirus (P < 0.05, N = 8 rats for each group) 2, 4, 6, and 9 days after infection (Figure 3C). These results suggest that overexpression of PKIA-AS1 alone is sufficient to produce hyperalgesia.

To investigate the effects of PKIA-AS1 on neuroinflammation associated with NP, we evaluated the activation of astrocyte and microglia as well as the expression levels of inflammatory cytokines interleukin (IL)-6, IL-1β, IL-12 and tumor necrosis factor (TNF-)-α expression in spinal cord. As expected, SNL induced astrocytic and microglial activation as evaluated by increased GFAP and Iba-1 expression levels, respectively, while PKIA-AS1 knockdown substantially reduced glial activation (Figure 4A). PKIA-AS1 knockdown also inhibited SNL-mediated upregulation of IL-6, IL-1β, and TNF-α at both mRNA and protein levels in spinal cord (Figure 4B,C). These findings suggest that knockdown of PKIA-AS1 represses neuropathic pain by inhibiting neuroinflammation.

The potential mechanisms underlying regulation of neuropathic pain by PKIA-AS1 were then investigated. To identify possible targets of PKIA-AS1, we screened numerous differentially expressed proteins in spinal cord tissues after PKIA-AS1 overexpression (Figure 5A). The detail information on the top 10 most upregulated and downregulated proteins is listed in Table 2. One of the most upregulated proteins is CDK6, which has been strongly implicated in inflammation (Handschick et al., 2014). We then confirmed the expression of CDK6 in spinal cord tissues from rats infected with PKIA-AS1. Indeed, Lv-PKIA-AS1 infection induced CDK6 overexpression at both mRNA and protein levels (Figure 5B,C). We further demonstrated that Lv-PKIA-AS1 infection upregulated CDK6 gene promotor activity (Figure 5D). We next tested whether PKIA-AS1 interacts with CDK6 promoter by conducting RNA pulldown assays in PC12 cells infected with Lv-PKIA-AS1 or control lentivirus. Both PKIA-AS1 and CDK6 promoter were more abundant in the Ago2 pellet than in the IgG pellet (Figure 5E), demonstrating that the CDK6 promoter interacts PKIA-AS1. We further tested how PKIA-AS1 regulates CDK6 promoter activity. The methylation status of CDK6 promoter was evaluated by quantitative methylation-specific PCR (qMSP) of the PC12 cells following Lv-PKIA-AS1 infection. Overexpression of PKIA-AS1 downregulated methylation of CDK6 promoter, suggesting that enhanced CDK6 expression in the SNL spinal cord is mediated by promoter hypomethylation induced by PKIA-AS1 (Figure 5F). In addition, PKIA-AS1 overexpression significantly inhibited DNMT1 expression, but not DNMT3A and DNMT3B expression (Figure 5G), suggesting that PKIA-AS1 regulates CDK6 expression by decreasing DNMT1-catalyzed methylation of CDK6 promoter.

We further investigated whether PKIA-AS1 controls neuropathic pain via CDK6. Expression of CDK6 was significantly increased in SNL model rats compared with control rats, whereas CDK6 expression was inhibited by PKIA-AS1 knockdown. In addition, PKIA-AS1 knockdown-mediated inhibition of CDK6 was reversed by Lv-CDK6 infection (Figure 6A,B). We found that Lv-CDK6 infection reversed the attenuation of mechanical allodynia and thermal hyperalgesia by Lv-shPKIA-AS1 infection following SNL (Figure 6C,D). Moreover, Lv-CDK6 reversed the suppression of IL-1β, IL-6, IL-12, and TNF-α mRNA and protein expression by Lv-shPKIA-AS1 (Figure 7A,B). Collectively, these results indicate that an epigenetic regulatory pathway involving PKIA-AS1 and CDK6 mediates SNL-induced neuropathic pain.