Water-Soluble Polymer Assists N-Methyl-D-Aspartic Acid Receptor 2B siRNA Delivery to Relieve Chronic Inflammatory Pain In Vitro and In Vivo

We constructed a water-soluble lipopolymer (WSLP) as a nonviral gene carrier to deliver siRNA targeting NR2B. The cytotoxicity and serum stability of WSLP loaded with siRNA were evaluated, and the knockdown efficiency of WSLP/NR2B-siRNA in PC12 cells was examined. The results showed that WSLP could protect the loading siRNAs from enzymatic degradation in serum and exhibit low cytotoxicity to cells. After transfection, WSLP/NR2B-siRNA complexes reduced the NR2B transcriptional level by 50% and protein level by 55% compared to control siRNA. Moreover, 3 days after intrathecal injection of WSLP/NR2B-siRNA complexes into rats, the NR2B protein expression decreased significantly to 58%, compared to control treatment (p < 0.01). Injection of WSLP with scrambled siRNA or of polyethylenimine (PEI) with NR2B-siRNA did not show this inhibitory effect. Additionally, injection of WSLP/NR2B-siRNA complexes significantly relieved inflammatory pain in rats at 3, 4, and 5 days with reduced MWT and decreased TWL scores, while injection of WSLP with scrambled siRNA or of PEI with NR2B-siRNA did not. These results demonstrated that WSLP can efficiently deliver siRNA targeting NR2B to PC12 cells and relieve pain in rats with chronic inflammatory pain.


Introduction
e phrase "neuropathic pain" appeared in public view only in the last decade and has been increasingly appreciated as a frequent source of chronic pain. Neuropathic pain is de ned as "pain caused by a lesion or disease of the somatosensory system." Nerve injury, nerve compression, diabetes, infection, and autoimmune disease may result in neuropathic pain [1].
Overexpression of N-methyl-D-aspartic acid receptor 2B (NR2B) plays an important role in the development of hyperalgesia [2]. Pain treatment research has focused on using siRNA targeting NR2B to suppress the expression of NR as an e ective way to treat chronic pain [3]. Finding an appropriate transduction tool is critical for the success of siRNA delivery into the human body. To avoid potential safety issues, nonvirus carriers are typically used for in vivo genetic treatment. However, commonly used means of nonvirus carrier treatment, such as liposome-mediated transfection and electroporation, have a desirable e ect in vitro but a reduced e ect in vivo [4].
RNA interference (RNAi) o ers great potential not only for in vitro target validation but also as a novel therapeutic strategy based on the highly speci c and e cient silencing of a target gene. Since it relies on small interfering RNAs (siRNAs), which are the mediators of RNAi-induced speci c mRNA degradation, a major issue is the delivery of therapeutically active siRNAs into the target tissue/target cells in vivo. For safety reasons, strategies based on (viral) vector delivery may be of only limited clinical use. e more desirable approach is to directly apply catalytically active siRNAs. Viral vectors are e cient gene transduction tools for in vitro cells. In vivo, nonviral vectors delivering siRNA have been used previously [5,6]. Polyethylenimine (PEI) is an organic material that can be used as a nonviral vector. Unfortunately, high-molecular-weight PEI is cytotoxic. Low molecular weight PEI has low cytotoxicity but also has a low transfection e ciency [7][8][9]. Recently, cholesterol has been used to modify low molecular weight PEI, and a water-soluble lipopolymer (WSLP) comprising cholesterol and low molecular weight PEI exhibited good biocompatibility, high transfection e ciency, and minimal toxicity [10][11][12][13]. A previous study indicated that WSLP can permeate the blood-brain barrier [14]. erefore, the present study hypothesizes that WSLP delivering siRNA targeting NR2B may inhibit NR2B expression in the spinal dorsal horn and may be used as a novel method for treating neuropathic pain. is WSLP exhibited high transfection e ciency and negligible toxicity, which owned the potential to target the central nervous system (CNS) in vivo [15][16][17][18]. erefore, we hypothesized that WSLP could be a reliable carrier for delivering NR2B-siRNA, which would suppress NR2B expression and consequently relieve chronic pain. Given that NR2B was the main cause of chronic in ammatory pain, in the present study, we attempted to treat the pain by knocking down the expression of NR2B using the WSLP delivery system described above. We prepared a siRNA targeting NR2B and fused it to WSLP, de ned as WSLP/NR2B-siRNA. Our data showed that WSLP/NR2B-siRNA eciently delivered siRNA into PC12 cells. In addition, we evaluated the MWT and TWD scores of the rat model with in ammatory pain treated with intrathecal injection of WSLP/NR2B-siRNA complexes in vivo.

Materials and Methods
2.1. WSLP Synthesis. WSLP was synthesized according to previously described methods [19]. Brie y, 30 mL of dehydrated chloroform (Guangzhou Chemical Reagent Factory, Guangzhou, China) was added to 100 μL of dehydrated trimethylamine and then mixed with 3 g of PEI (molecular weight ∼18,000, Suzhou Jingchun, Suzhou, China). is mixture was placed on ice, stirred for 30 min, mixed with 1 g of cholesterol chloroformate (Sigma-Aldrich Co. LLC, USA), dissolved in 10 mL of chloroform at 0°C, and stirred overnight. A yellow, viscous gel was obtained from this process, and it was dissolved in 50 mL of hydrochloric acid (0.1 mol/L). is dissolved gel was mixed with 200 mL of dehydrated chloroform to remove other polymers and cholesterol, and then, it was ltered through lter membranes (membrane pore size of 30 μm). e abovementioned procedures were repeated twice. e products were freeze-dried under low pressure to yield a yellow powder.

Stability of WSLP/NR2B-siRNA in Serum.
WSLP and NR2B-siRNA were mixed at a mass ratio of 5 : 1; equal amounts of the complexes were randomly placed into ve centrifugation tubes, and the samples were incubated at room temperature for 30 min. e samples were then infused with 10% aborted calf blood serum and incubated at 37°C for 0, 0.5, 1, 2, and 3 h. Each sample was subsequently electrophoresed as described above for 20 min, and the results were observed using a Gel Doc 2000 gel image analysis system (Bio-Rad, Hercules, CA, USA).

Transfection of WSLP/NR2B
-siRNA to Silence the NR2B. PC12 cells were randomly divided into three groups: the negative transfection group (NT group), transfected with NR2B-siRNA; the control transfection group (CT group), transfected with WSLP/scRNA; and the WSLP transfection group (WT group), transfected with WSLP/NR2B-siRNA. PCR and western blot assays were used to detect the mRNA and protein expression levels of NR2B in cells. Rats of both sexes were bred and maintained in a speci c pathogen-free unit at room temperature with humidity regulated (21 ± 2°C; 55 ± 10%) under 12/12 h light/dark cycle with lights o at 19:30 h and no twilight period. Mice were housed in individually ventilated cages (IVCs) (Tecniplast Sealsafe 1284L) receiving 60 air changes per hour, at a stocking density of 4-5 rats per cage unless otherwise stated below (overall dimensions of caging (L × W × H): 398 × 215 × 187 mm, oor area � 530 cm 2 ). Aspen bedding substrate and standard environmental enrichment of nestlet, cardboard tunnel, and three wooden chew blocks were provided. Rats were given water and breeding diet (irradiated A03, SAFE, France) ad libitum unless otherwise stated.
About 100 rats were randomly assigned to ve groups (20 rats per group): the control group (NS group), in which the rats received no treatment; the chronic in ammatory pain model group (CFA group), in which the rats had inammatory pain induced by intraplantar injection of complete Freund's adjuvant (CFA); the PEI/NR2B-siRNA group (PEI group), in which the rats were subjected to the same chronic in ammatory pain model but also received an intrathecal injection of low molecular weight PEI/NR2B-siRNA; the WSLP/NR2B-siRNA group (WSLP group), in which the rats were subjected to the same chronic inammatory pain model but also received an intrathecal injection of WSLP/NR2B-siRNA; and the WSLP/scRNA group (sWSLP group), in which the rats were subjected to the same chronic in ammatory pain model but also received an intrathecal injection of WSLP/scRNA. is study received permission from the Animal Ethics Committee of the Guangzhou General Hospital of the Guangzhou Military Command of the Chinese PLA, China.

Model Establishment and Intrathecal Injection of WSLP/siRNA.
e in ammatory pain model was established by injecting 120 µL of CFA (1 g/L) into the footpad of the rats. e NS group was administered an intraplantar injection of 120 μL of 0.9% normal saline instead of CFA. Chronic in ammatory pain rat models were done by subcutaneous injection of complete Freund's adjuvant (CFA) at the right rear toe. Brie y, the rats were anesthetized using 4% chloral hydrate (1 ml/100 g i.p.) and then subcutaneous injection of 120 μL CFA in the pain model rats and PBS in control models. e exclusion criteria were used to remove failed model rats as follows: the values of TWL and MWT were measured at day 1 after modeling, and the values decreased signi cantly proved successful modeling and these rats were included; vice verse, they were excluded from the experiments.
Various drugs were intrathecally administered according to previously described methods [20]. e drugs were slowly injected using a microsyringe vertically aligned to the L5-6 space. Successful intrathecal punctures were demonstrated by rats exhibiting a tail tremble or sudden lateral swing. One day after model establishment, the rats in the PEI, WSLP, and sWSLP groups received a single intrathecal injection of 20 μl of PEI/siRNA, WSLP/siRNA, or WSLP/scRNA, respectively. e mechanical withdrawal threshold (MWT) and thermal withdrawal latency (TWL) of 10 rats from each group were measured in the morning 1 day before model establishment as well as 3, 4, and 5 days after intrathecal injection. e other 10 rats from each group were intraperitoneally injected with 4% chloral hydrate 3 days after the intrathecal injection, and their spinal dorsal horns at L4 were subsequently harvested on ice and divided equally for PCR and western blot assays.
e PCR products were electrophoresed in an 1.5% agarose gel, and the electrophoresed gel was analyzed using a Gel Doc 2000 gel image analysis system (Bio-Rad, Hercules, CA, USA). e ratio of NR2B absorbance to β-actin absorbance was used to quantify the expression.

Western
Blot. Seventy-two hours after transfection, the incubation media were discarded, and the cells were washed 2-3 times with precooled PBS. e cells were then placed in microcentrifugation tubes, 400 μL of radio immunoprecipitation assay (RIPA) lysate and 4 μL of phenylmethylsulfonyl uoride (PMSF) were added to each tube on ice, and the cells were lysed on dry ice for 30 min. Finally, the cells were centrifuged at 12,000 ×g for 30 min at 4°C, and the resulting supernatants were stored at −80°C.
Spinal dorsal horns at L4-6 were ground into precooled tissue lysates at 4°C, placed in an ice bath for 5 minutes, and centrifuged at 800 ×g for 15 min. e supernatants were harvested, and their protein concentrations were measured using the bicinchoninic acid assay method as previously described [21]. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (with a separation gel concentration of 4% and a stacking gel concentration of 6%) was performed with 80 μg of samples in each lane. e proteins were semidried, transferred to polyvinylidene di uoride membranes, and blocked with 5% nonfat milk powder. e membranes were incubated with a goat anti-rat NR2B monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) in hybridization solution (Santa Cruz Biotechnology, Santa Cruz, Pain Research and Management 3 CA, USA) (1 : 500) for 2 h, followed by incubation with horseradish peroxidase-labeled rabbit anti-goat IgG hybridization solution (1 : 5000) for 1 h. e products were subjected to enhanced chemiluminescence and imaging with ultraviolet light. e target protein bands were analyzed using a Gel Doc 2000 gel image analysis system (Bio-Rad). e product of absorbance and area (OD × mm 2 ) was used to quantify protein expression.

Detection of Pain Behaviors.
MWT was determined using a Von Frey stimulation device (Von Frey Kit Cat. no. 1277, Ugo Basile Biological Research Apparatus, Varese, Italy). MWT detection was performed between 09:00 and 12: 00 in a quiet environment. 30 min prior to MWT detection, each rat was placed in a transparent organic glass container to adapt to the environment. Von Frey cilia of di erent sizes were used to vertically stimulate the sole of the hind foot for 6-8 s. Escape reactions included elevating the hind foot, delaying in putting down the hind foot, trembling, licking the hind foot, or escaping. e 50% withdrawal threshold was calculated [22].
TWL was determined using a plantar test (Plantar Test Cat. no. 37,370, Ugo Basile Biological Research Apparatus, Varese, Italy). e stimulus intensity was controlled at a range where the withdrawal reaction occurred at 12-15 s following the stimulation. e thermal stimulation lasted for 30 s. e interval between stimulations was 5 min. e detection was repeated in triplicate, and the mean value was calculated.
2.11. Statistical Analysis. All statistical analyses were performed with Statistical Package for the Social Sciences (SPSS) 13.0 software. e data are presented as the mean ± SD from three separate experiments. Statistical signi cance was determined by the paired or unpaired Student's t-test in cases of standardized expression data. Di erences were considered statistically signi cant at p < 0.05. We have added this paragraph in the Method section.

Feasibility and Stability of Polyelectrolyte Complex
Formed by WSLP with siRNA. To optimize the ratio between WSLP and siRNA in a complex, we determined the enwrapped siRNA by using electrophoresis. WSLP was conjugated with NR2B-siRNA at di erent mass ratios ranging from 1 : 1 to 7 : 1 and subjected the samples to electrophoresis. As shown in Figure 1(a), the brightness of the siRNA bands gradually faded with the increase of WSLP/siRNA ratios from 1 : 1 to 7 : 1. At a ratio of 5 : 1, the siRNA band was as weak as most of the siRNA in 6 : 1 and 7 : 1. Hence, we prepared the WSLP/siRNA complex at a ratio of 5 : 1 for future analysis (Figure 1(a)). en, the stability of the complex in serum was evaluated, as it was previously reported that WSLP conjugation would protect nucleic acid from degradation. Based on the serum stability of PEI-siRNA, control was previously conrmed several times by other researchers [23], and we did not repeat the examination. e naked siRNA was degraded in serum within 1.5 h, while the WSPL-conjugated siRNA could stay stable as long as 3 h in serum (Figure 1(b)). ese results suggested that the WSLP/siRNA could be used as a stable siRNA delivery tool.
In addition, the mRNA and protein levels of NR2B were further detected by qRT-PCR and WB, respectively. e results showed that there were no obvious alterations between the NS and sWSLP groups in NR2B expressions, while NR2B expressions in mRNA and protein levels in the PEI and WSLP groups were notably decreased as compared to the NS and sWSLP groups (Figures 1(c) and 1(d)).

Cytotoxicity of WSLP/siRNA In Vitro.
As a useful siRNA delivery tool, low cytotoxicity to cells is equally important for high e ciency. Next, we aimed to assess the cytotoxicity of WSLP/siRNA in vitro. We performed an MTS assay on untransfected PC12 cells (control group) and PC12 cells transfected with PEI-siRNA and various proportions of WSLP/NR2B-siRNA (1 : 1, 2 : 1, 3 : 1, 5 : 1, 10 : 1, and 20 : 1). We found that the resulting ODs of the PC12 cells transfected with any of the tested proportions of WSLP/NR2B-siRNA were not signi cantly di erent (p > 0.05) from those of PC12 cells transfected with PEI-siRNA and untransfected PC12 cells (Figure 2), con rming that WSLP/NR2B-siRNA complexes are not cytotoxic in these proportions.

NR2B Silencing by WSLP/siRNA Complexes In Vitro.
To con rm that NR2B-siRNA is able to e ectively silence NR2B when it is complexed with WSLP, we transfected PC12 cells with NR2B-siRNA (NT group), WSLP/scRNA (CT group), or WSLP/NR2B-siRNA (WT group) and assessed their NR2B mRNA and protein levels with PCR and western blot assays, respectively. e PCR results showed that the NR2B mRNA level of the WT group was signi cantly lower than that of the NT group, with a silencing rate of 50% (p < 0.01). e western blot results similarly showed that the average NR2B protein expression level of the WT group was 55% lower than that of the NT group (p < 0.01). ere was no signi cant di erence in the NR2B mRNA level or protein expression level between the NT and CT groups (p > 0.05; Table 1).
ese results show that WSLP/NR2B-siRNA silences NR2B more e ectively than NR2B-siRNA that is not complexed with WSLP.

NR2B Silencing by WSLP/siRNA Complexes In
Vivo. After nding that the WSLP/NR2B-siRNA complexes could knock down NR2B expression in vitro, we tested if they could also silence NR2B expression in vivo by using a rat model of chronic in ammatory pain.
We found that compared to the control NS group, the CFA, PEI, WSLP, and sWSLP groups showed signi cantly higher NR2B protein levels 3 days after administration of the intrathecal injection (p < 0.01; Figure 3(a)). Additionally, compared to the CFA group, the WSLP group showed 58% lower NR2B protein levels (p < 0.01), but the NR2B expression in the PEI and sWSLP groups was similar to that in the control group (p > 0.05; Figure 3(b)). In addition, the expression pattern of NR2B mRNA levels at 3 days after administration of the intrathecal injection in di erent groups was similar to its protein levels (Figure 3(c)).

Intrathecal Injection of WSLP/NR2B-siRNA Relieved
In ammatory Pain in Rats. We assessed the MWT and TWL of the rats, which re ect the degree of in ammatory pain of rats. e detailed timeline of animal model establishment was shown in Figure 4(a). Speci cally, signi cantly increased MWT and TWL indicated less pain. We found that the MWT (Figure 4(b)) and the TWL (Figure 4(c)) were signi cantly reduced in the a ected feet of the CFA group rats compared with the NS group rats at 4 days after model establishment via CFA injection (3 days after intrathecal injection) (p < 0.01). Moreover, these changes were maintained for 5 days after the intrathecal injection. e increase in MWT and TWL compared with that in the CFA group was the greatest in the WSLP group (p < 0.01). e MWT and TWL values for the PEI and sWSLP groups were similar to those for the CFA group (p > 0.05). ese ndings indicated that intrathecal injection of WSLP/ NR2B-siRNA relieved in ammatory pain in our rat model.

Discussion
e key to successful siRNA transfection in vivo is nding an appropriate vector with nontoxicity and high e ciency. Aigner et al. [24] believed that the most viable way is to directly link siRNA with the right vector. Recent lines of evidence discovered that WSLP has shown promise as a genetic carrier with high e ciency and low toxicity [13]. Moreover, WSLP consisted of polyethylenimine (PEI1800) and cholesterol. PEI1800 was also called low molecular weight polyethylenimine, which had lower cytotoxicity and transfection e ciency, compared with PEI25000 (highmolecular-weight polyethylenimine). Previous studies have  proved that the biological characters of PEI with di erent molecular weights are di erent [25]. (1) PEI of high molecular weight (∼25,000) exhibited better transfection e ciency than PEI of low molecular weight (<5000) [26][27][28]; (2) PEI of high molecular weight (∼25,000) had higher toxicity to cells, and PEIs can be hampered by their cellular toxicity in vivo. e results in our study were consistent with this, as the PEI synthesized here (∼1800) exhibited low toxicity to cells but not satis ed the transfection e ect in vivo. rough modi cation, PEI (1800) was conjugated with cholesterol to construct the WSLP, and the e ciency was signi cantly improved.
Of note, the transfection e ciency signi cantly increased when PEI1800 conjugated to cholesterol [29]. Consistently, in our study, the result showed that WSLP/siRNA has a high stability in serum. NR2B expressions at mRNA and protein levels in the PEI and WSLP groups were lower than those in the NS and sWSLP groups. e cytotoxicity test results showed that, at mass ratios of WSLP/siRNA 1 : 1-7 : 1, the ODs obtained and observed in the MTS tests were not signi cantly di erent between WSLP/NR2B-siRNA-transfected and nontransfected cells. ese results suggested that WSLP has a low cytotoxicity to cells and a good stability in serum, which entitled favorable value in application.
As we all know, the ratio of carrier and siRNA played a key role in the siRNA transfection, which can signi cantly a ect the transfection e cacy [30]. Zhang et al. [31] found that when the ratio of G5 PD dendrimer to siRNA increased, the e cacy of transfection was upregulated and the expression of target gene was decreased, and 3.5 : 1 was the optimal ratio for silencing the gene expression. In the present study, we synthesized WSLP based on the method mentioned by Lee for example, [16] and discovered that this polymer can e ectively bind NR2B-siRNA when they were mixed at a mass ratio of 5 : 1. We then tested the stability of WSLP/siRNA complexes in serum and found that WSLP/NR2B-siRNA still produced a visible electrophoresis band after incubation in serum for 3 h, indicating that it was not degraded. Previous reports have proved that inhibiting NR2B gene expression contributed to chronic in ammatory pain relief. In this study, we investigated whether WSLP could deliver NR2B-siRNA to successfully inhibit NR2B gene expression in cells and consequently reduce chronic in ammatory pain in the rat model. We found that WSLP/NR2B-siRNA complexes reduced the NR2B transcriptional level and the protein level compared with unmodi ed siRNA in PC12 cells.
ere were no signi cant di erences between the NR2B mRNA or protein expression levels of the sWSLP and untransfected groups.
ese results indicated that WSLP could deliver siRNA to e ectively inhibit the genetic expression of NR2B in PC12 cells.
Intrathecal injection was an e ective route of administration to avoid recognition and elimination of drugs for gene therapy by the reticuloendothelial system in the blood [32,33]. erefore, we aimed to observe the inuence of WSLP/NR2B-siRNA on NR2B gene expression and its therapeutic e ect on chronic in ammatory pain. e results showed that NR2B protein expression was e ciently inhibited by intrathecal injection of WSLP/NR2B-siRNA complexes; the protein level was reduced compared with that in untreated rats with chronic in ammatory pain. In contrast, injection of WSLP complexed with control scRNA or of PEI with NR2B-siRNA did not produce this inhibitory e ect, indicating that the intrathecal injection of WSLP/NR2B-siRNA complexes can speci cally inhibit NR2B gene expression in rats with chronic in ammatory pain. us, WSLP can e ciently deliver siRNA targeting NR2B in vivo to inhibit NR2B gene expression in the spinal dorsal horn. In addition, the MWT and TWL scores were signi cantly decreased at each time point in the CFA-injected paw compared with those in normal rat paws. Intrathecal injection of WSLP/NR2B-siRNA complexes lessened the MWT and TWL reduction caused by CFA injection, while the intrathecal injection of PEI/siRNA and WSLP/scRNA did not exhibit such an e ect. us, the intrathecal injection of WSLP/NR2B-siRNA complexes relieved chronic in ammatory pain in rats.

Conclusion
It is the rst study to con rm that WSLP can e ciently deliver siRNA targeting NR2B to inhibit NR2B gene expression in the spinal dorsal horn and relieve chronic in ammatory pain in rats. It has been reported that intrathecal injection of NR2B-siRNA can relieve nociception of chronic in ammatory pain in rats but is unable to relieve the hyperalgesia of rats with chronic neuropathic pain. siRNA targeting the gene subunit is easily degraded by enzymes and is unable to enter cells, which restricts the therapeutic e ect [32,34]. Wu treated chronic inammatory pain in rats using adenovirus vector-mediated siRNA to silence NRs [35]. However, the heterologous defect of viral vectors remains an unsolved concern. For example, Garraway et al. [36] found that heterologous HIV-siv vector had a defect in the transduction of dendritic cells (DCs) and macrophages. Although further studies on WSLP/NR2B-siRNA are needed to con rm its use as a nontoxic and high-e ciency vector, our study provides a novel approach to gene therapy, targeting NRs for the treatment of chronic in ammatory pain.