Adult male Sprague-Dawley rats (n=163; 200–220 g; 8–10-weeks-old) were purchased from the Experimental Animal Center of Guangdong Province (Guangdong, China). Rats were housed in a temperature-controlled room (24±2°C) with a 12 h light/dark cycle and food and water available ad libitum. The animals were allowed to acclimate for ≥3 days, prior to the experiments approved by the Animal Use and Care Committee for Research and Education of The First People's Hospital of Foshan (Guangdong, China), in accordance with the ethical guidelines of the International Association for the Study of Pain (25). Pain and suffering were minimized, as was the number of animals used.

The CCI neuropathic pain model was established as previously proposed (26). The animals were anesthetized by intraperitoneal injection of 10% chloral hydrate (400 mg/kg). The biceps femoris was dissected to expose the left sciatic nerve. The area proximal to the sciatic nerve's trifurcation was freed of adhering tissue, and 4-0 surgical catgut suture was tied loosely around it with approximately 1 mm between ligatures. Sham-operated rats underwent the same surgery, but without ligation. Following the surgery, muscle and skin tissue was sutured in layers, and treatment with ampicillin (100 mg/kg) intraperitoneally for the first day was administered. Rats in the naïve group received neither surgery nor ligation. Each group contains 6 rats. Rats in the CCI group exhibiting no mechanical and thermal allodynia at 7 days post-surgery were excluded from the study.

In order to determine the effect of single intrathecal apelin-13 (10 µg; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) and ML221 (1, 3, 10 and 30 µg; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) doses on mechanical allodynia and heat hyperalgesia, the rats were treated with light isoflurane anesthesia (1–2%). The injection was carried out with a 30-gauge needle as previously described (27). Apelin-13 or ML221 was slowly administered in a solution of 10 µl.

A permanent intrathecal catheter (PE-10 polyethylene tube; BD Biosciences, Franklin Lakes, NJ) was inserted between the T₃ and T₄ vertebrae and extended to the L₄ and L₅ segments following intraperitoneal administration of 400 mg/kg chloral hydrate, in order to assess the effect of continuous injection of apelin-13 or ML221 using behavioral testing. Cannulated animals were housed individually; those with surgery-associated neurological ailments were not taken into account in the final analysis. Drugs were injected through the catheter in a volume of 10 µl followed by flushing with 10 µl saline. All injections were performed by the same investigator.

Mechanical allodynia was evaluated as previously described (28). Each rat was placed individually into a wire mesh-bottomed cage for 30 min of adaptation prior to the test. Filaments were presented, from weakest to strongest, at a 90° angle to the plantar surface, with enough force to slightly bend the paw, and maintained for 6 sec. A positive response was indicated by the animal quickly withdrawing or flinching the paw. The stimuli were presented at intervals of 2 min, and the stimuli were increased and decreased sequentially (‘up-and-down’ technique) to assess paw-withdrawal threshold (PWT). Interpolation of the 50% threshold (the weakest force causing ≥3 withdrawals in 5 successive applications) was carried out according to the method of Dixon (29). The experiments were performed in a blinded manner.

Paw-withdrawal latency (PWL) values following thermal hyperalgesia were determined using a Plantar Analgesia Meter (IITC Life Science, Woodland Hills, CA, USA) (30). Individual rats were assessed three times (left hind paw) and average values were obtained.

The total RNA was obtained from L₄-L₅ spinal cord segments using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's protocol. RNA purity and concentration were evaluated using the NanoDrop 2000 (Thermo Fisher Scientific, Inc., Wilmington, DE, USA). RT-qPCR analysis was carried out to determine the mRNA concentration of apelin and APJ using the EXPRESS One-Step SYBR Green ER SuperMix kit (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's protocol, on a StepOne Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). Relative gene expression was assessed by the 2^−ΔΔCq method (31). RT-qPCR amplification was carried out as previously described (30), with β-actin as the reference gene for normalization. The experiments were performed in triplicate. The specific primers used for sequence detection were as follows: β-actin, forward 5′-GGAGATTACTGCCCTGGCTCCTA-3′ and reverse 5′-GACTCATCGTACTCCTGCTTGCTG-3′; Apelin, forward 5′-CTCTCCTTGACTGCCGTGTGTG-3′ and reverse 5′-GCATGTTGCCTTCTTCTAGCCC-3′; APJ, forward 5′-TGCCTCAACCCCTTCCTCTA-3′ and reverse 5′-GTTCTCCTCCCTTGCACATG-3′.

In order to evaluate the variation in the protein concentrations of the apelin-APJ system (rats randomly divided into control group, sham group and CCI groups) and ERK signaling [rats randomly divided into sham group, CCI group, CCI + dimethylsufoxide (DMSO) group and CCI + ML221 group], western blot analysis was performed with L₄-L₅ spinal cord segment specimens. Sample homogenization was performed in lysis buffer as described previously (30). Protein amounts were determined using the bicinchoninic acid assay. Equal amounts (20 µg) of total protein were resolved by 10% SDS-PAGE and electro-transferred onto polyvinylidene fluoride membranes (Merck KGaA). The membranes were blocked using 5% nonfat dry milk for 1 h at room temperature (25°C) and incubated overnight at 4°C with primary antibodies raised against apelin (goat; cat. no. sc-33469; 1:1,500), APJ (rabbit; cat. no. sc-33823; 1:2,000), ERK1/2 (mouse; cat. no. sc-514302; 1:1,500), and p-ERK1/2 (rabbit; cat. no. sc-101760; 1:1,000), all purchased from Santa Cruz Biotechnology, Inc. For loading control, the blots were probed with GAPDH antibody (mouse; cat. no. sc-32233; 1:5,000; Santa Cruz Biotechnology, Inc.). These membranes were further incubated for 2 h with rabbit anti-goat (cat. no. BA1060; 1:2,000; Wuhan Boster Biological Technology, Ltd., Wuhan, China), goat anti-mouse (cat. no. BA1051; 1:2,000; Wuhan Boster Biological Technology, Ltd.) and mouse anti-rabbit (cat. no. L27A9; 1:2,000; Cell Signaling Technology, Inc., Danvers, MA, USA) IgG horseradish peroxidase-conjugated secondary antibodies at room temperature. Band intensities were assessed using Image J software (version 1.48u; National Institutes of Health, Bethesda, MD, USA).

When neuropathic pain was fully established, the rats were intraperitoneally administered 400 mg/kg chloral hydrate for deep anesthesia, and L₄-L₅ spinal cord segments were extracted, fixed in 4% paraformaldehyde for 90 min at room temperature (25°C) and then placed in 30% sucrose solution at 4°C overnight, embedded in optimal cutting temperature (OCT) compound (Thermo Fisher Scientific, Inc.), and sectioned at 5 µm for immunofluorescence. Blocking with 2% donkey serum (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature was followed by overnight incubation (4°C) with primary antibodies raised against apelin (goat; cat. no. sc-33469; 1:100; Santa Cruz Biotechnology, Inc.) and APJ (rabbit; cat. no. sc-33823; 1:100; Santa Cruz Biotechnology, Inc.). DyLight 649- (rabbit anti-goat; cat. no. E032630; 1:200; EarthOx Life Sciences, Millbrae, CA, USA) or Andy Fluor 488-(goat anti-rabbit; cat. no. L110B; 1:300; GeneCopoeia, Inc., Rockville, MD, USA) conjugated secondary antibodies were added for 1 h at room temperature. Immunoreactivity was captured on a confocal laser scanning microscope (magnification, ×200) and analyzed with Adobe Photoshop CS6 (Adobe Systems, Inc., San Jose, CA, USA).

All values are expressed as the mean ± standard error of the mean. The data on mechanical allodynia and thermal hyperalgesia were analyzed by two-way repeated measures analysis of variance followed by the Bonferroni post hoc test. Western blotting and RT-qPCR data were analyzed by one-way analysis of variance followed by the Newman-Keuls post hoc test. Independent samples Student's t-test was used to analyze the immunofluorescence data. P<0.05 was considered to indicate a statistically significant difference.

The function of the apelin-APJ system in chronic neuropathic pain was investigated by examining apelin-APJ system expression at the mRNA and protein level in CCI rats. CCI caused rapid and persistent neuropathic pain from the third day following surgery, which lasted for >2 weeks in the rats. mRNA and protein levels were assessed at 3, 7 and 14 days post-injury; these time points corresponded, respectively, to the onset, peak and full establishment of allodynia post-CCI. RT-qPCR data demonstrated that apelin mRNA levels continuously increased, peaking at 7 days post-CCI, with a 3-fold increase in the CCI group compared with the controls (Fig. 1A). In addition, a parallel, significant and time-dependent increase in APJ mRNA levels, was observed at 3, 7 and 14 days post-CCI (Fig. 1B). Western blot analysis demonstrated that CCI caused a rapid (within 3 days) and long-lasting (>14 days) increase in APJ protein amounts in the spinal cord (Fig. 2B). Additionally, spinal apelin protein levels were significantly increased, compared with sham-operated animals, 7 and 14 days post-surgery (Fig. 2A).

In order to examine spinal cord apelin and APJ expression levels, immunostaining was carried out in naïve and CCI-operated mice. As the expression levels of apelin and APJ peaked at 7 days post-surgery, this time point was selected to assess the alteration of apelin and APJ expression levels in the spinal cord. As presented in Fig. 3B, APJ was constitutively expressed in the superficial dorsal horn. CCI induced a marked increase in APJ expression and an increased number of APJ-immunoreactive cells was observed in the CCI group compared with the sham group (Fig. 3B). In addition, low immunofluorescence intensities were obtained for apelin in sham animals, while CCI rats exhibited enhanced expression 7 days post-surgery (Fig. 3A).

The results of the present study suggested that pain-associated alterations of apelin and APJ levels in the superficial dorsal horn may be involved in the development and maintenance of chronic neuropathic pain.

Mechanical allodynia and thermal hyperalgesia in the animals were measured at 1 day before (baseline), and 1, 3, 5 and 7 days following, CCI or sham surgery. It was identified that CCI led to a gradual reduction in PWT and PWL to a level considered to be consistent with neuropathic pain, in accordance with a previous report (26).

In order to investigate the role of increased expression of the spinal apelin-APJ system in pain sensory modulation, intrathecal injections of apelin-13 and ML221 were performed in CCI rats. A single intrathecal injection of 10 µg ML221, 7 days following CCI, caused a transient, marked reduction of mechanical allodynia and heat hyperalgesia compared with CCI rats treated with normal saline (Fig. 4A and B). By contrast, an intrathecal injection of apelin-13 did not affect mechanical allodynia or heat hyperalgesia (Fig. 4A and B), suggesting that the spinal apelin-APJ system contributed to CCI-induced neuropathic pain. Drug dosages were selected based on a previous study (20).

In order to study the dose-effect association of ML221 in alleviating CCI-induced neuropathic pain, increasing doses of ML221 (1, 3, 10 and 30 µg) were injected intrathecally 7 days post-surgery. As presented in Fig. 4C and D, ML221 inhibited mechanical allodynia and thermal hyperalgesia in a dose-dependent manner; these effects began 1 h post-injection, peaked at 2 h and diminished within 4 h. A dose of 3 µg ML221 transiently reduced CCI-induced mechanical allodynia and heat hyperalgesia, although the reduction was not statistically significant; this result was not observed with a 1 µg dose. A 10 µg dose of ML221 significantly increased PWT and PWL between 1 and 4 h. Increased ML221 dosages displayed analgesic effects similar to those obtained with 10 µg, indicating that a single intrathecal injection of 10 µg ML221 may be considered the optimal dose for the reversal of allodynia.

In order to assess the long-term impact of chronic intrathecal ML221 administration during the initiation of CCI-induced neuropathic pain, 10 µg ML221 was administered once daily for 3 days (days 0–2). Pretreatment with ML221 for 3 successive days markedly delayed CCI-induced mechanical allodynia and heat hyperalgesia for 2–11 days (Fig. 5A and B). In order to determine the long-term effect of repeated treatment with ML221 in maintaining CCI-induced neuropathic pain, the same ML221 doses were given at days 7–9, and produced a somewhat reliable inhibitory effect on allodynia (Fig. 5C and D). Analgesia persisted for 4 days following the third treatment with ML221. The present results suggested that a reduction in the expression of spinal APJ may inhibit the development and maintenance of CCI-induced neuropathic pain.

As presented above, increased expression of the spinal apelin-APJ system contributes to CCI-induced neuropathic pain. As ERK signaling is implicated in the effects of apelin, the present study investigated whether apelin-induced nociceptive behaviors were mediated via the ERK signaling pathway. Rats were assigned to sham, CCI, CCI + DMSO, and CCI + ML221 groups, respectively. A single intrathecal injection of ML221 (10 µg) or vehicle (DMSO) was performed 7 days post-surgery, and L₄-L₅ SC segments were harvested 2 h post-injection for the assessment of ERK and phosphorylated ERK levels. As presented in Fig. 6, phosphorylated ERK levels were increased in the CCI group compared with the sham group. Phosphorylated ERK induction was alleviated by a single intrathecal injection of ML221, corroborating the results obtained in the behavioral tests as described above. The present data indicated that the spinal apelin-APJ system may be involved in neuropathic pain via the ERK signaling pathway.