DKK3 ameliorates neuropathic pain via inhibiting ASK-1/JNK/p-38-mediated microglia polarization and neuroinflammation

Male Sprague–Dawley rats (200–240 g) were supplied by Tongji hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. The rats were housed under controlled conditions (temperature: 22–25℃, relative humidity: 45–65%, and 12-h light to dark cycle, with food and water ad libitum). All experiments were approved by the Experimental Animal Care and Use Committee of Tongji hospital, Tongji Medical College, Huazhong University of Science and Technology. All experiments were conducted in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals, and the ARRIVE Guidelines for Reporting Animal Research.

A rat model of neuropathic pain was induced by SNI as described previously [36‐38]. Briefly, the rats were anesthetized with 2.5% isoflurane, the right sciatic nerve and its three branches (the common peroneal, tibial, sural nerves) were exposed. Then, the common peroneal and tibial nerves were ligated. The distal to the ligation was sectioned, cutting 2 to 4 mm of the distal nerve stump. For the sham-surgery rats, the sciatic nerve was exposed without ligation.

To measure mechanical allodynia, paw withdrawal threshold (PWT) in response to von Frey filament stimuli was assessed as described previously [39]. In brief, the rats were placed in individual plastic boxes on a metal mesh floor and allowed to habituate for 30 min. The von Frey filaments (1, 1.4, 2, 4, 6, 8, 10, and 15 g) were applied for up to 6 s per filament to the mid-plantar of the right hind paw. The positive response was defined as the sudden paw withdrawal, licking, and shaking. When a positive response was appeared, the paw was retested after a 5-min rest, starting with the next descending von Frey filament. When no response occurred, the next higher filament was applied. The lowest amount of force required to elicit a positive response was recorded as the PWT (in grams). All of the behavioral tests were performed by an investigator who was blinded to the experimental design.

The OFT was used for locomotor activity measurement, which was performed as previously described [14, 40]. Briefly, motor activity was tested in an open field (60 × 60 × 30 cm). Each rat was placed in the center of the open field and was allowed to move freely for 5 min. A video camera was located 100–120 cm above the floor for behavior recording. Following the completion of a test, 75% alcohol was used to clean the arena. Locomotor behavior was analyzed by counting the total distance and average speed.

As described previously [39], intrathecal (i.t.) catheters was carried out 5 days prior to the establishment of SNI models. Briefly, the rats were anesthetized using isoflurane (2.5%). Then, the PE10 polyethylene catheters (inner diameter 0.3 mm, outer diameter 0.6 mm) were inserted from L5-L6 spinous processes. The correct position of the catheter was verified by a tail flick response immediately after inserting the catheter and further confirmed using an i.t. injection of 2% lidocaine. Animals exhibiting motor dysfunction were excluded from the experiments. No animal was excluded in this study.

rDKK3 (SRP6268, Sigma-Aldrich, MO, USA) was dissolved in saline. The choice of solvent is based on previous study [32]. During administration with rDkK3, we prepared fresh rDKK3 solution every day and gave it to animals within 15 min. In this study, we aimed to investigate a purely spinal mechanism. Therefore we chose intrathecal injection and the dosage of rDKK3 was determined by our preliminary experiments. To determine whether a single dose of rDKK3 could attenuate established mechanical allodynia in SNI rats, rDKK3 (10, 30, or 50 μg, i.t.) was given on day 7 after surgery. The behavioral test was conducted before rDKK3 injection, and 1, 2, 4, and 6 h after the injection. To determine whether repeated injection of rDKK3 could reverse mechanical allodynia in SNI rats, rDKK3 (30 μg, i.t.) was given once daily for five consecutive days starting from day 7. The behavioral test was performed on day 6 and 2 h after rDKK3 injection each day. To determine whether early treatment with rDKK3 could suppress the development of mechanical allodynia in SNI rats, rDKK3 (30 μg, i.t.) was given once daily for five consecutive days starting from day 1 after the surgery. The behavioral test was conducted before the operation and 2 h after rDKK3 injection on day 3, 7, 8, 9, 10 and 14 after SNI. To identify whether Kremen-1 and DVL-1 are involved in DKK3 regulating microglia polarization and alleviating neuropathic pain in SNI rats, rDKK3 (30 μg, i.t.) was given once daily for five consecutive days starting from day 7 after the operation. The behavioral test was conducted before the operation, and on day 3, 5, 7, 8, 9, 10 and 11 after SNI.

siRNAs were synthesized by Tsingke Biotechnology (Beijing, China). siRNA duplexes that specifically targeted Kremen-1 was: sense 5’-CCUCUCGCAUCCAUUUCAATT-3’, and anti-sense 5’-UUGAAAUGGAUGCGAGAGGTT-3’. siRNA duplexes that specifically targeted DVL-1 was: 5’-CCGAGAUGGAAUGGACAAUTT-3’, and anti-sense 5’-AUUGUCCAUUCCAUCUCGTT-3’. Scramble siRNA was synthesized by a scrambled sequence of nucleotides as a control siRNA, scramble siRNA was: sense 5’-UUCUCCGAACGUGUCACGUTT-3’, anti-sense 5’-ACGUGACACGUUCGGAGAATT-3’. The siRNA was dissolved in RNase-free water at 1 μg/μl and mixed with the transfection reagent branched polyethyleneimine (PEI; Sigma-Aldrich) and 5% glucose for 10 min at room temperature before use. The Kremen-1 siRNA, DVL-1 siRNA or scramble siRNA was administrated via i.t. in 1, 3, 5, 7 days after nerve injury.

Rats were euthanized with CO₂, the L4–L6 spinal cord segments were quickly excised and homogenized in phosphate-buffered saline (PBS). The supernatant was collected by centrifugation at 15,000g at 4 °C for 60 min, and was analyzed using rat IL-1β (Cat#: 88-6010-22; Invitrogen; Carlsbad, CA, USA), TNF-α (Cat#: 88-7340-22; Invitrogen), and IL-6 (Cat#: 88-50625-22; Invitrogen) ELISA kits, according to the manufacturer’s instructions.

Rats were euthanized with CO₂, the L4–L6 spinal cord segments were quickly excised. Then, the spinal cord was homogenized in an ice-cold mixture of radioimmunoprecipitation assay lysis buffer, phosphatase inhibitor, and phenylmethylsulfonyl fluoride (Boster Biological Technology, Wuhan, Hubei, China), and then centrifuged at 12,000 rpm at 4 ℃ for 30 min. The supernatants were collected, and the protein concentration was determined using a BCA protein assay kit (Cat#: AR0146, Boster Biological Technology). The proteins were boiled at 95 °C in a loading buffer for 10 min. Equivalent amounts of samples (30 μg protein) were separated using 10% sodium dodecyl sulfate–polyacrylamide gels electrophoresis and transferred onto polyvinylidene fluoride membranes (Cat# IPVH00010; Millipore, Billerica, MA, USA). After blocking with 5% bovine serum albumin in Tris-buffered saline and Tween 20 (0.1%) (TBST) for 2 h at room temperature (RT), the membranes were incubated overnight at 4 ℃ with rabbit anti-β-actin antibody (1:5000, Cat# AC026; ABclonal, Woburn, MA, USA); rabbit anti-DKK3 antibody (1:1000, Cat# ab186409; Abcam, Cambridge, UK); rabbit anti Kremen-1 antibody (1:1000, Cat# ab211285; Abcam); rabbit anti DVL-1 antibody (1:1000, Cat# ab233003; Abcam); rabbit anti-ASK1 antibody (1:1000; Cat# A3271; ABclonal); rabbit anti-p-ASK1(1:1000; Cat# 3764; Cell Signaling Technology, Danvers, MA, USA); rabbit anti-JNK antibody (1:1000; Cat# 9258; Cell Signaling Technology); rabbit anti-p-JNK (1:1000; Cat# 4668; Cell Signaling Technology); rabbit anti-p38 antibody (1:1000; Cat# A5049; ABclonal); rabbit anti-p-p38 antibody (1:1000; Cat# 4511; Cell Signaling Technology); goat anti-ionizedcalcium-binding adapter molecule 1(Iba-1, microglial marker) antibody (1:500; Cat# ab5076; Abcam); rabbit anti-CD16 antibody (1:1000; Cat# ab211151; Abcam); rabbit anti-CD86 antibody (1:1000; Cat# ab112490; Abcam); rabbit anti-iNOS antibody (1:1000; Cat#18985-1-AP; Proteintech, Chicago, IL, USA); rabbit anti-Arg1 antibody (1:1000; Cat# A4923; ABclonal); rabbit anti-CD206 antibody (1:500; Cat# ab64693; Abcam); rabbit anti-IL-10 antibody (1:500; Cat# A12255; ABclonal); rabbit anti-IL-1β antibody (1:1000; Cat# AF4006; Affinity Bioscinence; OH; USA); rabbit anti-TNF-α antibody (1:500; Cat# ab205587; Abcam); rabbit anti-IL-6 antibody (1:500; Cat# A0286; ABclonal). The membranes were then washed in TBST and incubated with horseradish peroxidase-conjugated goat anti-rabbit antibody (1:5000; Cat# A21020; Abbkine, Wuhan, Hubei, China), or donkey anti-goat antibody (1:5000; Cat# AS031; ABclonal) for 2 h at RT. The bands were finally visualized with SuperLumia ECL Plus HRP Substrate Kit (Cat# K22030; Abbkine) and then detected using a computerized image analysis system (Bio-Rad, ChemiDoc XRS + , USA). The intensity of protein blot was quantified using System with Image Lab software (Bio-Rad Laboratories), normalized to loading control β-actin antibody and expressed as the fold of control. The blot density of sham, sham + Vehicle, or scramble siRNA groups was set as 1.

As described previously [41], to isolate microglia, spinal cord tissues were removed and minced with scissors in ice‐cold Dulbecco’s modified Eagle medium (DMEM) (Invitrogen). Thereafter, a monocular suspension was formed by digesting with 0.25% trypsin (Invitrogen) in a water bath at 37 °C for 30 min, and then were co-stained for CD86-FITC (an M1 microglia biomarker; 1:200; Cat# 555018; BD Biosciences, San Jose, CA, USA) and CD206-APC (an M2 microglia biomarker; 1:200; Cat# 550889; BD Biosciences) for 45 min at room temperature following the manufacturer’s instructions. The samples were detected using NovoCyte 2040R (ACEA Biosciences) and then analyzed by FlowJo software v.7.6.1.

Under deep anesthesia with isoflurane (2.5%), the rats were perfused intracardially with 0.1 M PBS, followed by 4% ice-cold paraformaldehyde (PFA) in PBS. Then the L4-L6 spinal cord was removed and post-fixed in 4% PFA for 4 h, and subsequently dehydrated in 30% sucrose solution overnight at 4 ℃. The collected spinal cord samples were sectioned to 20 μm thickness in a cryostat (CM1900, Leica, Wetzlar, Germany). For single immunofluorescent staining, the sections were blocked with 5% donkey serum and 0.3% Triton X-100 for 1 h at RT, and then incubated with rabbit anti-DKK3 antibody (1:50, Cat# ab186409; Abcam); rabbit anti Kremen-1 antibody (1:50, Cat# ab211285; Abcam); rabbit anti DVL-1 antibody (1:50, Cat# ab233003; Abcam); rabbit anti-p-ASK1(1:50; Cat# 3764; Cell Signaling Technology); goat anti-Iba-1 (1:100; Cat# ab5076; Abcam) overnight at 4 ℃. After washing 3 times in PBS, the sections were incubated with CoraLite594-conjugated donkey anti-rabbit secondary antibody (1:100; Cat# SA00013-8; Proteintech) or FITC-conjugated affinipure donkey anti-goat secondary antibody (1:50; Cat# SA00003-3; Proteintech) for 2 h at RT and stained with DAPI for 5 min. For double-immunofluorescence, the sections were blocked with 5% donkey serum and 0.3% Triton X-100 for 1 h at RT, and then incubated with a mixture of rabbit anti-DKK3 antibody (1:50, Cat# ab186409; Abcam), or rabbit anti Kremen-1 antibody (1:50, Cat# ab211285; Abcam), or rabbit anti DVL-1 antibody (1:50, Cat# ab233003; Abcam), or rabbit anti-p-ASK1 antibody (1:50; Cat# 3764; Cell Signaling Technology), or rabbit anti p-JNK antibody (1:50; Cat# ab131499; Abcam), and mouse anti-GFAP antibody (1:100; Cat# 3670; Cell Signaling Technology) or mouse anti-neuronal nuclei antibody (NeuN, neuronal marker) antibody (1:50; Cat# ab104224; Abcam) or goat anti-Iba-1antibody (1:50; Cat# ab5076; Abcam) overnight at 4 ℃. After washing 3 times in PBS, the sections were incubated with a mixture of CoraLite594-conjugated donkey anti-rabbit secondary antibody (1:100; Cat# SA00013-8; Proteintech) and CoraLite488-conjugated donkey anti-mouse secondary antibody (1:50; Cat# SA00013-5; Proteintech) or FITC-conjugated affinipure donkey anti-goat secondary antibody (1:50; Cat# SA00003-3; Proteintech) for 2 h at RT and stained with DAPI for 5 min. Images were captured using a fluorescence microscope (BX51, OLYMPUS, Japan). As described previously [42‐45], immunohistochemistry analysis was calculated by using ImageJ software (National Institutes of Health, Bethesda, MD, USA) and averaged per mouse, and each group included 6 animals. The numbers of DKK3, Kremen-1, DVL-1, p-ASK1, Iba-1-positive cells in the spinal dorsal horn were counted in a 500 μm × 500 μm measuring frame. We counted every third sections (40 μm apart) to avoid counting the same cell in more than one section. The cell counts were then used to determine the total number of positive cells per square millimeter. Cells double-labeled for DKK3, or Kremen-1, or DVL-1, or p-ASK1, or p-JNK, and the cell-specific markers NeuN, Iba-1 and GFAP were quantified too in sham and SNI 7d groups. We recorded the number of cells double-labeled with DKK3, or Kremen-1, or DVL-1, or p-ASK1, or p-JNK, and a cell-specific marker in a 500 μm × 500 μm measuring frame in the spinal dorsal horn. The cell counts were then used to determine the total number of double-positive cells per square millimeter.

In brief, total RNAs were isolated with RNA isolator total RNA extraction reagent (Vazyme, R401-01) according to the manufacturer’s instructions. cDNA was synthesized using PrimeScipt RT Master Mix (Takara, RR036A). Ten microliters of qPCR reactions were prepared from 5 μl Premix Ex TaqII (Takara, RR820A), 0.5 μl primer (final concentration 10 nM), 2 μl DEPC water, and 2 μl cDNA. Reactions were run in a LightCycler 480 qPCR instrument (Roche) using the standard conditions 95 °C for 5 min, 40 cycles (95 °C for 15 s, 56 °C for 30 s, and 72 °C for 30 s) plus melting curve. Relative levels were quantified with the 2-ΔΔCT method that was normalized to the sham group. The primers used were as follows: β-actin forward: 5’-CCCATCTATGAGGGTTACGC-3’, β-actin reverse: 5’-TTTAATGTCACGCACGATTTC-3’; DKK3 forward: 5’-CCCCGACGGCCACTTGGACTC-3’, DKK3 reverse: 5’-GCCGCTTCTTCCGCCTCCATCT-3’; Kremen1 forward: 5’-CGGGCACCAGTAAGACGTCCAACA-3’, Kremen1 reverse: 5’-TGCCTCCCCGTGCTTCCAGTAGTC-3’; DVL-1 forward: 5’-TCGGGGTGGTGAAGGAGGAGATCT-3’, DVL-1 reverse: 5’-CCCCAATGCCGCCTGTCCTCTC-3’.

In the present study, all rats were randomly assigned to the following three separate experiments which are shown in the timeline of experimental design. The experimental groups and number of animals used in experiments are listed in Fig. 1.

As shown in Fig. 2a, there is no significant difference regarding the ipsilateral PWT among sham and SNI rats at baseline. However, the ipsilateral PWT in SNI rats was markedly decreased from day 3 to day 14. In contrast, as shown in Fig. 2b, the contralateral PWT in sham and SNI rats had no significant change during the observation period. These results indicate that SNI successfully induced the development of mechanical allodynia. To explore the potential role of DKK3, Kremen-1 and DVL-1 in neuropathic pain, we examined DKK3, Kremen-1 and DVL-1 expression in spinal cord on day 3, 7, and 14 following SNI. As shown in Fig. 2c–e, DKK3, Kremen-1 and DVL-1mRNA levels were persistently reduced on day 3, 7, and 14 after SNI, when compared with sham control. Western blot revealed that the protein levels of DKK3, Kremen-1 and DVL-1 were significantly decreased on day 3 and sustained on day 14 (Fig. 2f–h). Furthermore, single immunofluorescence further showed SNI-induced lower expression of DKK3, Kremen-1 and DVL-1 in the spinal dorsal horn at 3, 7, and 14 days after SNI, but not in sham control (Fig. 2i–n).

To define the type of cells expressing DKK3, Kremen-1 and DVL-1 in the spinal dorsal horn, we performed co-staining of DKK3, Kremen-1 and DVL-1with neuronal marker (NeuN), microglial marker (Iba-1) and astrocytic marker (GFAP). As shown in Fig. 3a–f, in sham-operated rats, DKK3, Kremen-1 and DVL-1 immunoreactivity (IR) was co-stained excessively with NeuN and Iba1, to a lesser extent, with GFAP. We also determined DKK3, Kremen-1 and DVL-1cellular localization at day 7 after SNI, the results demonstrated that the ratio of DKK3, Kremen-1 and DVL-1 co-localization with NeuN or Iba-1 was decreased in SNI 7d group compared with the Sham group (Fig. 3g–i).

Then we examined the protein expression levels of ASK-1, p-ASK-1, JNK, p-JNK, p38, p-p38 in spinal cord. As shown in Fig. 4a, b, single immunofluorescence showed SNI-induced higher expression of p-ASK1 in the spinal dorsal horn at 3, 7, and 14 days after SNI, but not in sham control. To define the type of cells expressing p-ASK1 in the spinal dorsal horn, we performed co-staining of p-ASK1 with neuronal marker (NeuN), microglial marker (Iba-1), and astrocytic marker (GFAP) in sham and SNI 7d group. As shown in Fig. 4c–e, double-immunofluorescence revealed that ratio of p-ASK1 co-localization with Iba-1 was significantly increased in SNI 7d group compared with Sham group. Moreover, it was found that there was no significant difference in ASK-1 (Fig. 4f), JNK (Fig. 4h), p38 (Fig. 4j) protein expression between SNI 3d, 7d, 14d and sham-operated rats. However, the results showed that the protein expression level of p-ASK1 (Fig. 4g), p-JNK (Fig. 4i) and p-p38 (Fig. 4k) were significantly elevated on day 3 and sustained on day 14 after SNI compared with the sham group. Besides, to define the type of cells expressing p-JNK in the spinal dorsal horn, we performed co-staining of p-JNK with NeuN, Iba-1, and GFAP in the sham and SNI 7d group. As shown in Additional file 1: Fig. S1a-e, double-immunofluorescence revealed that ratio of p-JNK co-localization with Iba-1, GFAP was significantly increased in SNI 7d group.

Then, we determined whether rDKK3 could attenuate mechanical allodynia in neuropathic pain rats. To determine whether a single dose of rDKK3 could attenuate established mechanical allodynia in SNI rats, rDKK3 (10, 30, or 50 μg, i.t.) was given on day 7 after surgery. The behavioral test was conducted before rDKK3 injection, and 1, 2, 4, 6 h after the injection. As shown in Fig. 5a, there is no significant change regarding ipsilateral PWT in SNI rats treated with 10 μg ZLN005 compared with SNI + Vehicle group. However, rDKK3 (30 and 50 μg) significantly increased the ipsilateral PWT in SNI rats, peaking at 2 h, and lasted for at least 4 h compared with SNI + Vehicle group. To determine whether repeated injection of rDKK3 could reverse mechanical allodynia in SNI rats, rDKK3 (30 μg, i.t.) was given once daily for five consecutive days starting from day 7. The behavioral test was performed on day 6 and 2 h after rDKK3 injection each day. As shown in Fig. 5c, repetitive injections of rDKK3 (30 μg, i.t.) considerably reversed established mechanical allodynia in SNI rats. To determine whether early treatment with rDKK3 could suppress the development of mechanical allodynia in SNI rats, rDKK3 (30 μg, i.t.) was given once daily for five consecutive days starting from day 1 after the surgery. The behavioral test was conducted before the surgery, and on day 3, 7, 8, 9, 10 and 14 after surgery. As shown in Fig. 5e, the ipsilateral PWT was significantly increased from day 3 to day 8 in rDKK3-treated SNI rats compared with vehicle-treated SNI rats. Moreover, rDKK3 did not affect the PWT on the contralateral side (Fig. 5b, d, f). These results indicate that rDKK3 can markedly attenuate established mechanical allodynia in SNI rats, and also delayed the development of mechanical allodynia in SNI rats. Based on the results from the OFT, there was no significant difference in the total distance and average speed between the sham and SNI rats treated with Vehicle or rDKK3 (Fig. 5g–i). The OFT result indicated that intrathecal injection with rDKK3 did not affect the motor function of rats.

Then, we tested whether treatment with rDKK3 could inhibit the activated ASK-1/JNK/p38 MAPK signaling pathway caused by SNI. Firstly, treatment with rDKK3 attenuated the reduced protein level of DKK3 caused by neuropathic pain (Fig. 6a). Moreover, as shown in Fig. 6b, western blot results displayed that the protein expression level of ASK-1 had no remarkable difference in sham and SNI rats applied with Vehicle or rDKK3. However, rDKK3 could significantly alleviate the increased level of p-ASK-1 in the spinal cord induced by nerve injury compared with the SNI + Vehicle group (Fig. 6c). The results also indicated that there was no significant difference in the protein expression of JNK between sham and SNI rats administrated with Vehicle or rDKK3 (Fig. 6d). But compared with the SNI + Vehicle group, application with rDKK3 dampened the up-regulated levels of p-JNK in the spinal cord of rats with SNI (Fig. 6e). Then, it was found that the level of p38 had no prominent difference in sham and SNI rats treated with Vehicle or rDKK3 (Fig. 6f). Nevertheless, rDKK3 reversed the raised level of p-p38 in the spinal cord of rats with SNI (Fig. 6g).

As microglial activation plays an important role in neuropathic pain, we sought to identify whether the increasing spinal DKK3 would alter the status of microglia in spinal dorsal horn after SNI. Microglial activation was demonstrated by the changes of number and morphology in Iba1-positive cells in lamina I–III layers of the dorsal horn. Firstly, we evaluated microglial markers Iba-1 protein expression in the spinal cord, western blot results showed that SNI robustly up-regulated the level of Iba1, and which could be abolished by administration with rDKK3 (Fig. 7a). Moreover, as shown in Fig. 7b, c, SNI triggered radical changes in the morphology and accumulation of Iba1 stained microglia in the spinal dorsal horn. By contrast, SNI induced these alterations of microglia were significantly attenuated by intrathecal injection of rDKK3. Furthermore, we found that rDKK3 prevented SNI-induced M1 microglial polarization and promoted the M2 phenotype. Western blot results showed that application with rDKK3 decreased the protein expression level of M1 markers including CD16 (Fig. 7d), CD86 (Fig. 7e), iNOS (Fig. 7f) and increased the protein expression level of M2 markers including Arg1 (Fig. 7g), CD206 (Fig. 7h), IL-10 (Fig. 7i) in the spinal cord of rats with SNI. Moreover, flow cytometry further confirmed that rDKK3 promoted the switch of microglia from M1 type to M2 type in rats with neuropathic pain (Additional file 2: Fig. S2a-b). Besides, we also observed that rDKK3 could ameliorate neuroinflammation induced by neuropathic pain. Western blot results showed that rDKK3 could reverse the up-regulated levels of pro-inflammatory cytokines including IL-1β (Additional file 3: Fig. S3a), IL-6 (Additional file 3: Fig. S3b), and TNF-α (Additional file 3: Fig. S3c) in the spinal cord. Furthermore, ELISA results further indicated that rDKK3 could alleviate the neuroinflammation induced by neuropathic pain (Additional file 3: Fig. S3d-f).

To verify whether Kremen-1 and DVL-1 were participated in the analgesic effect of DKK3, the SNI rats were given rDKK3 in addition to either Kremen-1 siRNA, DVL-1 siRNA, or scramble siRNA. The knockout effect of siRNA was detected by western blot in 11 days following SNI. As shown in Fig. 8a, b, the results showed that Kremen-1 siRNA and DVL-1 siRNA resulted in significant downregulation of Kremen-1 and DVL-1 protein expression. Furthermore, as shown in Fig. 8c, the ipsilateral PWT results indicated that administration with rDKK3 alleviated mechanical allodynia induced by nerve injury compared with SNI + Vehicle group, but this effect was abolished by Kremen-1 siRNA or DVL-1 siRNA. Moreover, Kremen-1 siRNA, DVL-1 siRNA and scramble siRNA all did not affect the PWT on the contralateral side (Fig. 8d). According to the OFT result, there was no significant difference in the total distance and average speed between the Sham group, SNI + Vehicle group, SNI + rDKK3 group, SNI + rDKK3 + scramble siRNA group, SNI + rDKK3 + Kremen-1 siRNA group, and SNI + rDKK3 + DVL-1 siRNA group (Fig. 8e–g). The OFT result demonstrated that intrathecal injection with Kremen-1 siRNA or DVL-1 siRNA did not affect the motor function of rats.

Then we evaluated the effect of siRNA on ASK-1/JNK/p38 MAPK pathway. As shown in Fig. 9a, western blot results indicated that there was no significantly difference in the protein expression level of ASK1 between the Sham group, SNI + Vehicle group, SNI + rDKK3 group, SNI + rDKK3 + scramble siRNA group, SNI + rDKK3 + Kremen-1 siRNA group, and SNI + rDKK3 + DVL-1 siRNA group. However, application with rDKK3 reduced the increased protein expression level of p-ASK1 in the spinal cord of rats with SNI, while the effect was reversed by intrathecal injection with Kremen-1 siRNA or DVL-1 siRNA (Fig. 9b). Furthermore, the results showed that there was no significantly difference in the protein expression level of JNK between the Sham group, SNI + Vehicle group, SNI + rDKK3 group, SNI + rDKK3 + scramble siRNA group, SNI + rDKK3 + Kremen-1 siRNA group, and SNI + rDKK3 + DVL-1 siRNA group (Fig. 9c). However, treatment with rDKK3 ameliorated the increased level of p-JNK in the spinal cord of rats with SNI, while Kremen-1 siRNA or DVL-1 siRNA dampened the effect (Fig. 9d). Moreover, western blot results implied that there was no significantly difference in the expression level of p38 between the Sham group, SNI + Vehicle group, SNI + rDKK3 group, SNI + rDKK3 + scramble siRNA group, SNI + rDKK3 + Kremen-1 siRNA group, and SNI + rDKK3 + DVL-1 siRNA group (Fig. 9e). But, administration with rDKK3 down-regulated the raised level of p-p38 in the spinal cord of rats with SNI, while Kremen-1 siRNA or DVL-1 siRNA reversed the effect (Fig. 9f).

Firstly, we evaluated microglial markers Iba-1 protein expression in the spinal cord, western blot results showed that Kremen-1 siRNA and DVL-1 siRNA both could reverse the reduced protein level of Iba-1 caused by rDKK3 (Fig. 10a). Moreover, as shown in Fig. 10b-c, SNI triggered radical changes in the morphology and accumulation of Iba-1 stained microglia in the spinal dorsal horn. SNI induced these alterations of microglia were significantly abolished by intrathecal injection of rDKK3, while administration with Kremen-1 siRNA and DVL-1 siRNA both could abolished the effect of rDKK3 on microglia activation. Moreover, western blot results showed that rDKK3 down-regulated the increased levels of CD16 (Fig. 10d), CD86 (Fig. 10e), iNOS (Fig. 10f) and up-regulated the decreased levels of Arg1 (Fig. 10g), CD206 (Fig. 10h), IL-10 (Fig. 10i) compared with the SNI + Vehicle group, while both knock out Kremen-1 with Kremen-1 siRNA and knock out DVL-1 with DVL-1 siRNA could reverse the effect of rDKK3 on microglia polarization. Flow cytometry further identified that rDKK3 promoted the switch of microglia from M1 type to M2 type in rats with neuropathic pain, while Kremen-1 siRNA and DVL-1 siRNA could abrogate the effect of rDKK3 on microglia polarization (Additional file 4: Fig. S4a-b). In addition to this, western blot (Additional file 5: Fig. S5a-c) and ELISA (Additional file 5: Fig. S5d-f) results showed that Kremen-1 siRNA and DVL-1 siRNA both could reverse the attenuated neuroinflammation in the spinal cord induced by rDKK3 compared with the SNI + rDKK3 group.