EphrinBs/EphBs Signaling Is Involved in Modulation of Spinal Nociceptive Processing through a Mitogen-activated Protein Kinases-dependent Mechanism

• ❖ Eph receptor tyrosine kinases and their ligands, ephrins in sensory neurons may play a role in hypersensitivity after nerve injury

• ❖ In the periphery, Eph receptors stimulate protein kinase (MAPK) activity

• ❖ In mice, intrathecal injection of ephrins stimulated MAPK activity in spinal cord neurons and astrocytes and caused hypersensitivity

• ❖ Blockade of MAPK activity prevented hypersensitivity from ephrins and blockade of Eph receptors in the spinal cord reduced hypersensitivity and MAPK activity in a model of neuropathic pain

EPH receptor tyrosine kinases and their membrane-bound ligands, ephrins, are involved in diverse aspects of development, such as tissue patterning, angiogenesis, axon guidance, and synapse formation.1–4Recent advances indicate that Eph receptors and ephrin ligands are present in the adult brain and peripheral tissue and play a critical role in modulating multiple aspects of physiology and pathophysiology (e.g. , activity-dependent synaptic plasticity, regulation of pain threshold, epileptogenesis, inflammation response, and excitotoxic neuronal death).5–9Interestingly, several Eph receptors and ephrin ligands are also expressed in the adult rat spinal cord and the dorsal root ganglion.10–12Bundesen et al.  12reported that EphB2 receptor was present in the laminae I–III of the dorsal horn and on small- and medium-sized dorsal root ganglion neurons but not on large-diameter neurons, two important sites for modulation of nociceptive information. The interactions of ephrinB–EphB modulate synaptic efficacy in the spinal cord, contributing to sensory abnormalities in persistent pain states in an N -methyl-d-aspartate (NMDA) receptor- dependent manner.13,14Recently, our and other studies also demonstrated that activation of spinal ephrinBs/EphBs system played a critical role in the development and maintenance of chronic pain after peripheral nerve injury.15–17These studies indicated that Ephrin/Eph system may be involved in physiologic and pathologic pain modulation in spinal cord level. However, the downstream mechanisms that control this process are not well understood.

It is well established that the mitogen-activated protein kinases (MAPKs) activation is involved in the modulation of nociceptive information and peripheral and central sensitization produced by intense noxious stimuli.18–23Several lines of evidence have shown that regulation of MAPKs is associated with EphBs receptors activation.24–27Our recent study showed that MAPKs signaling pathway mediated hyperalgesia induced by the activation of peripheral ephrinBs/EphBs system.28Thus, it is possible that MAPKs, as the downstream effectors, participate in modulation of nociceptive information related to ephrinBs/EphBs signaling in the spinal cord level. We provide strong evidence to support this hypothesis.

Adult male Kunming mice (20–25 g) were used in these studies. Mice were housed under a 12 h/12 h light–dark cycle regimen, with free access to food and water. The animals were provided by Experimental Animal Center of Xuzhou Medical College. All experimental protocols were approved by the Animal Care and Use Committee of Xuzhou Medical College (Xuzhou, Jiangsu Province, China) and were in accordance with the Declaration of the National Institutes of Health Guide for Care and Use of Laboratory Animals  (publication no. 80–23, revised 1996).

SB203580, a p38 kinase inhibitor; U0126, the mitogen-activated protein or extracellular signal-regulated kinase inhibitor; and SP600125, a Jun N-terminal kinases (JNK) inhibitor were purchased from Biomol Research Laboratories (Plymouth Meeting, PA). All MAPKs inhibitors were dissolved in 1% dimethyl sulfoxide (DMSO). EphrinB1-Fc and EphB1-Fc were purchased from R&D Systems Inc. (Minneapolis, MN) and are dissolved in saline. All doses of drugs are based on the results of preliminary experiments. The dose of each drug and time points of treatment were presented in the parts of figure legends.

The method described by Hylden and Wilcox29was used for an intrathecal injection of drugs. In brief, a 28-gauge stainless steel needle attached to a 25-μl Hamilton microsyringe was inserted between the L5 and L6 vertebrae in conscious mice. A sudden slight flick of the tail indicated entry into the subarachnoid space. A volume of 5 μl of drug solution or physiologic saline was injected over a 30-s period into the subarachnoid space, and the injection cannula was left in place for a further 15 s. Motor function was evaluated by observation of placing or stepping reflexes and righting reflexes at 2 min before nociceptive test. Animals with signs of motor dysfunction were excluded from the experiments.

Chronic constrictive injury (CCI) model was performed following the method of Bennett and Xie.30In brief, mice were anesthetized with sodium pentobarbital (40 mg/kg, intraperitoneal injection). Left sciatic nerve was exposed at mid-thigh level through a small incision, and a unilateral constriction injury just proximal to the trifurcation was performed with three loose ligatures using a 5-0 silk thread (spaced at a 1-mm interval). In sham-operated animals, the nerve was exposed but not ligated. The incision was closed in layers, and the wound was treated with antibiotics.

The procedure used was essentially the same as that reported by Hunskaar and Hole.31Approximately 30 min before testing, mice were individually placed in perspex observation chambers (10 × 20 × 15 cm) for adaptation. Then, the animals were taken out of the chamber, and 10 μl of 1% formalin in 0.9% saline was injected subcutaneously into the dorsal surface of the right hind paw with a 25-μl Hamilton syringe with a 28-gauge needle. Immediately after formalin injection, each mouse was returned to the observation chamber. The amount of time spent licking and biting the injected paw and the number of flinching paw were measured from 0 to 5 min (the first phase) and from 10 to 40 min (the second phase) after formalin injection and was considered as indicative of nociception.

Thermal hyperalgesia was measured using the paw-withdrawal latency (PWL) according to the method described by Hargreaves et al.  32In brief, mice were placed in clear plastic chambers (7 × 9 × 11 cm) and allowed to acclimatize to the environment for 1 h before testing. The radiant heat was directed to the plantar surface of each hind paw that was flushed against the glass or site of injection of solution through the glass plate. The nociceptive endpoints in the radiant heat test were the characteristic lifting or licking of the hind paw. The time to the endpoint was considered the PWL. The radiant heat intensity was adjusted to obtain basal PWL of 12–15 s. An automatic 20-s cutoff was used to prevent tissue damage. Thermal stimuli were delivered three times to each hind paw at 5-min intervals.

In CCI model, withdrawal latencies were measured in the same animal on both the ipsilateral (ligated) and the contralateral (unligated) paw. Results were expressed as the difference scores of PWL (s) = contralateral latency − ipsilateral latency.

Mechanical allodynia was assessed by using von Frey filaments (North Coast Medical, Inc., San Jose, CA), starting with 0.31 g and ending with 4.0 g filament as cutoff value. Animals were placed in individual plastic boxes (20 × 25 × 15 cm) on a metal mesh floor and allowed to acclimate for 1 h. The filaments were presented, in ascending order of strength, perpendicular to the plantar surface with sufficient force to cause slight bending against the paw and held for 6–8 s. Brisk withdrawal or paw flinching was considered as positive responses. The paw-withdrawal threshold (PWT) was determined by sequentially increasing and decreasing the stimulus strength (the “up-and-down” method), and the data were analyzed using the nonparametric method of Dixon, as described by Chaplan et al.  33 

In CCI model, withdrawal thresholds were measured in the same animal on both the ipsilateral (ligated) and the contralateral (unligated) paw. Results were expressed as the difference scores of PWT (g) = contralateral threshold − ipsilateral threshold. The behavioral testing was performed by an investigator blinded to the treatment.

Mice were anesthetized with sodium pentobarbital (60 mg/kg, intraperitoneal injection) and were subjected to sternotomy followed by intracardial perfusion with 20 ml saline and 100 ml 4% ice-cold paraformaldehyde in 0.1 mol/l phosphate buffer. The spinal cord of L^4–5was removed, postfixed in 4% paraformaldehyde for 3 h, and subsequently allowed to equilibrate in 30% sucrose in phosphate buffer overnight at 4°C. Thirty micrometer transverse series sections were cut on a cryostat and stored in phosphate buffer. After washing in phosphate buffer saline, the tissue sections were incubated in phosphate buffer saline containing 5% normal goat serum and 0.3% Triton X-100 at room temperature for 30 min. For the Fos protein assay, the sections were incubated in primary polyclonal rabbit-anti-Fos antibody (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C for 48 h. The sections were then incubated in biotinylated goat anti-rabbit (1:200) at 37°C for 1 h and in avidin–biotin–peroxidase complex (1:100; Vector Labs, Burlingame, CA) at 37°C for 2 h. Finally, the sections were treated with 0.05% diaminobenzidine for 5–10 min. Sections were rinsed in phosphate buffer saline to stop the reaction, mounted on gelatin-coated slides, air-dried, dehydrated with 70–100% alcohol, cleared with xylene, and cover slipped for microscopic examination. For analysis of the change of Fos protein expression, we examined five L^4–5spinal cord sections per animal, selecting the sections with the greatest number of positive neurons. For each animal, we recorded the total number of positive neurons in the bilateral spinal cord I–V and X lamina. All positive neurons were counted without considering the intensity of the staining.

For double immunofluorescence staining, the sections were incubated overnight at 4°C with a mixture of monoclonal antiphosphorylated MAPKs (p-MAPKs) antibody (anti-p-p38 from Signalway [Nanjing, China], anti-p-ERK and anti-p-c-JNK from Cell Signaling Technology [Danvers, MA], 1:200) and monoclonal neuronal-specific nuclear protein (NeuN; neuronal marker, antimouse, 1:100; Chemicon, Temecula, CA) or glial fibrillary acidic protein (astrocyte marker, anti-mouse, 1:200; Chemicon) followed by a mixture of Rhodamine Red-X goat anti-rabbit immunoglobulin G and fluorescein isothiocyanate monoclonal rat anti-mouse immunoglobulin G (1:100; Invitrogen, Carlsbad, CA) for 2 h at room temperature. Nonspecific staining was determined by excluding the primary antibodies. Sections were rinsed, mounted, and cover slipped with glycerol containing 2.5% of antifading agent 1,4-diazabicyclo[2.2.2]octane (Sigma, St. Louis, MO) and stored at −20°C in the dark. Images were captured using a laser scanning confocal microscopy (TCS SP2; Leica Microsystems Inc., Wetzlar, Germany).

The spinal cords of mice were quickly extracted and stored in liquid nitrogen. Tissue samples were homogenized in lysis buffer containing (in millimoles): Tris, 20.0 mm; sucrose, 250.0 mm; Na³VO⁴, 0.03 mm, MgCl², 2.0 mm, EDTA, 2.0 mm, EGTA, 2.0 mm; phenylmethylsulfonyl fluoride, 2.0 mm; dithiothreitol, 1.0 mm; protease inhibitor cocktail, 0.02% (v/v); pH 7.4. The homogenates were centrifuged at 5,000 g  for 30 min at 4°C. The supernatant was collected, and protein concentration was measured according to the Bradford34method, using bovine serum albumin as standard. The protein samples were stored at −80°C.

Protein samples were dissolved in 4× sample buffer (pH 6.8): Tris–HCl, 250.0 mm; sucrose, 200.0 mm; dithiothreitol, 300.0 mm; Coomassie brilliant blue-G, 0.01%; and sodium dodecyl sulfate, 8% and denatured at 95°C for 5 min, then the equivalent amounts of proteins (30 μg) were separated by using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto nitrocellulose membrane. In addition, the gels stained with Coomassie blue were used to confirm the equal amounts of protein loaded on each lane. The membranes were incubated overnight at 4°C with the following primary antibodies: primary polyclonal rabbit anti-p-p38 or anti-p38 antibody (1:400), primary polyclonal rabbit anti-p-ERK1/2 or anti-ERK1/2 antibody (1:400), or primary polyclonal rabbit-anti-p-JNK or anti-JNK antibody (1:400). The specificity for p-MAPKs antibodies was confirmed by loss of bands in the absence of primary antibodies. The membranes were extensively washed with Tris-buffered saline Tween 20 and incubated for 1 h with the secondary antibody conjugated with alkaline phosphatase (1:1,000) at room temperature. The immune complexes were detected by using a nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate assay kit (Sigma). Western blot densitometry analysis of signal intensity was performed using Adobe Photoshop software (Adobe, San Jose, CA), and phosphorylation levels of MAPKs from densitometry were normalized to total MAPKs. The blot density from control groups was set as 100%.

Data are expressed as mean ± SEM. Statistical analysis between two samples was performed using Student t  test. Statistical comparison of more than two groups was performed using one-way ANOVA followed by a Tukey post hoc  test. The significance of any differences in thermal latency and mechanical threshold in behavior test was assessed using two-way ANOVA. “Time” was treated as “within subjects” factor and “treatment” was treated as a “between subjects” factor. The area under the pain threshold change versus  time curve was calculated by GraphPAD Prism5 (Graph Pad Software, Inc., San Diego, CA) in some behavioral test. Statistical analyses of data were generated using GraphPAD Prism 5. All P  values given are based on two-tailed tests. A value of P  less than 0.05 was considered as statistically significant.

Previous study has shown that activation of spinal EphBs receptors by intrathecal injection of ephrinB2-Fc produced prolonged thermal hyperalgesia, but not mechanical allodynia, in the rats.13Our results were in partial agreement with this study. We found that intrathecal injection of ephrinB1-Fc (0.02, 0.1, and 0.5 μg in 5 μl saline), not control Fc (see experimental protocol in fig. 1A), induced a dose-dependent thermal hyperalgesia and mechanical allodynia in the mice, which can last at least up to 24 h and return to baseline level on 48 h after injection of ephrinB1-Fc (P < 0.05, from 0.5 to 24 h time point, ephrinB1-Fc0.1 or ephrinB1-Fc0.5 group vs.  saline or Fc group; fig. 1B). Compared with saline and Fc groups, the calculated area under the curve (−2–48 h) in PWL and PWT tests was significantly decreased in a dose-dependent manner in ephrinB1-Fc group (fig. 1Binset). To rule out a nonspecific effect through the Fc portion, human Fc was used as a control. No significant hyperalgesia was detected after injection of human Fc. Intrathecal injection of both ephrinB1-Fc and ephrinB2-Fc can activate EphBs receptor. To date, we cannot identify which EphBs subgroup was activated dominantly by ephrinB1-Fc or ephrinB2-Fc. Various species, not various regents, used in these two experiments may be the main cause of difference in mechanical allodynia induced by intrathecal injection of ephrinBs-Fc in rats and mice.

Spinal neuronal sensitization was involved in the development and maintenance of hyperalgesia induced by the different causes. Fos protein, the product of c-fos immediate early gene, has been used as a maker for neuronal activation in the central nervous system.35,36There is a positive correlation between the quantity of Fos protein expression and the neuronal activation induced by nociceptive stimuli in spinal cord neurons. To further clarify the algesic effect of intrathecal injection of ephrinB1-Fc, we investigated the change of spinal Fos protein expression induced by intrathecal injection of different doses of ephrinB1-Fc (0.02, 0.1, and 0.5 μg in 5 μl saline). Intrathecal injection of ephrinB1-Fc, not saline or Fc, induced a dose-dependent increase in spinal Fos protein expression. The expression of Fos protein mainly distributed in I–V lamina and surrounding of center canal (X lamina) of spinal cord (P < 0.0001, ephrinB1-Fc0.1 or ephrinB1-Fc0.5 group vs.  saline or Fc group; fig. 1C). These results further confirmed that activation of spinal ephrinBs/EphBs system can sensitize the spinal neurons and induce pain behaviors in naïve mice.

Some studies have shown that spinal MAPKs activation mediated inflammatory or injury-induced spinal sensitization and hyperalgesia.18–23Furthermore, studies in vitro  and in vivo  also have indicated that ephrinBs/EphBs signaling is involved in the regulation of MAPKs activity.24–27,37,38Thus, we ask whether activation of spinal MAPKs pathway contributes to ephrinB1-Fc-induced pain behaviors. Phosphorylation level of MAPKs correlates with MAPKs activation and is used routinely as an indicator of MAPKs activation.18In the current studies, MAPKs activation was evaluated by detecting the expression of p-MAPKs by Western blot analysis in mice. We first analyzed the time course of expression of spinal p-MAPKs, including p-p38, p-ERK, and p-JNK, after intrathecal injection of ephrinB1-Fc (0.5 μg). A rapid ERK activation was detected at 10 min, and the significant increase in p-p38 and p-JNK expression was detected at 30 min after intrathecal injection of ephrinB1-Fc. Activation of three MAPKs lasted at least up to 2 h and returned to baseline levels on 8 h after intrathecal injection of ephrinB1-Fc (p-p38: P < 0.01 at 0.5 h, P < 0.05 at 2 h; p-ERK: P < 0.01 at 0.17 h, P < 0.05 at 0.5 and 2 h; p-JNK: P < 0.05 at 0.5 and 2 h, vs.  0-h; fig. 2A). To further determine ephrinB1-induced MAPKs activation, we picked up 30-min time point to perform dose–response experiments. Intrathecal injection of ephrinB1-Fc (0.02, 0.1, and 0.5 μg), not Fc, induced a dose-dependent increase in spinal p-MAPKs expression (P < 0.05 or P < 0.01, ephrinB1-Fc0.1 or ephrinB1-Fc0.5 group vs.  Fc group; fig. 2B).

Activated MAPKs in both neurons and spinal glial cells were involved in spinal nociceptive information modulation.23,39–42Previous studies have shown that ephrinBs or EphBs is expressed in astrocytes of the adult spinal cord.12To identify the cell types that expressed spinal p-MAPKs after intrathecal injection of ephrinB1-Fc, double immunofluorescence staining of p-MAPKs was performed with cell-specific markers: NeuN for neurons, glial fibrillary acidic protein for astrocytes at 30 min after intrathecal injection of ephrinB1-Fc (0.5 μg; fig. 3). We found that the activated MAPKs by intrathecal injection of ephrinB1-Fc were primarily expressed in spinal cord dorsal horn (fig. 3A). Furthermore, MAPKs in dorsal root ganglion also were activated by intrathecal injection of ephrinB1-Fc (fig. 3C). All three phosphorylated MAPKs colocalized with the neuronal marker (NeuN; fig. 3B) and p-p38 and p-JNK, not p-ERK, with the astrocyte marker (glial fibrillary acidic protein) in the spinal cord (fig. 3D).

The current data have indicated that intrathecal injection of ephrinB1-Fc induced thermal hyperalgesia and mechanical allodynia, which was associated with activation of spinal MAPKs. Therefore, inhibition of activation of spinal MAPKs will alleviate pain behaviors in theory. As shown in figure 4, pretreatment with a p38 MAPK inhibitor SB203580 (SB), an MEK inhibitor U0126 (U) or SP600125 (SP), a JNK MAPK inhibitor (0.02, 0.1, and 0.5 μg in 1% DMSO for each inhibitor) at 30 min before ephrinB1-Fc injection (see experimental protocol in fig. 4A) dose dependently prevented thermal hyperalgesia (fig. 4C) and mechanical allodynia (fig. 4D) induced by intrathecal injection of ephrinB1-Fc (0.5 μg). Compared with DMSO-ephrinB1-Fc group, the calculated area under the curve (−2–48 h) in PWL and PWT tests was significantly increased in a dose-dependent manner in SB-ephrinB1-Fc, U-ephrinB1-Fc, and SP-ephrinB1-Fc groups (fig. 4C and Dinset). It suggested that activation of MAPKs was involved in the initiation of pain behaviors by ephrinB1-Fc. Then, we asked whether spinal MAPKs pathway also participate in its maintenance process. To address this question, SB, U, or SP (0.5 μg for each inhibitor) was administrated at 30 min after intrathecal injection of ephrinB1-Fc (0.5 μg; see experimental protocol in fig. 4B). We found that all three MAPKs inhibitors reversed the established thermal hyperalgesia and mechanical allodynia by intrathecal injection of ephrinB1-Fc. This effect lasted at least up to 8 h after injection of MAPKs inhibitors (fig. 4E). Compared with ephrinB1-Fc-DMSO group, the calculated area under the curve (−2–8 h) in PWL and PWT tests was significantly increased in ephrinB1-Fc-SB, ephrinB1-Fc-U, and ephrinB1-Fc-SP groups (fig. 4Einset).

The expression of Fos protein may also be a useful tool to examine the effectiveness of different analgesic regimens.35–36To further clarify the analgesic effect of inhibition of MAPKs on pain induced by intrathecal injection of ephrinB1-Fc, we investigated the effect of pretreatment or posttreatment with MAPKs inhibitors on spinal Fos protein expression induced by intrathecal injection of ephrinB1-Fc (0.5 μg). DMSO, SB (0.5 μg), U (0.5 μg), or SP (0.5 μg) was intrathecally administrated at 30 min before or after intrathecal injection of ephrinB1-Fc (0.5 μg), and spinal Fos protein expression was assayed at 2 h after ephrinB1-Fc injection. Both pre- and posttreatment with all three MAPKs inhibitors inhibited the spinal Fos expression induced by intrathecal injection of ephrinB1-Fc (P < 0.01; fig. 5).

Our behavioral and molecular findings revealed that intrathecal injection of ephrinB1-Fc activated all three MAPKs pathways; however, inhibition of any one of the three MAPKs pathways could have almost completely prevented or reversed pain behaviors and spinal Fos expression induced by intrathecal injection of ephrinB1-Fc. To confirm whether inhibition of any one of the three MAPKs also decreased phosphorylation of the two other MAPKs as a consequence, DMSO, SB (0.5 μg), U (0.5 μg), or SP (0.5 μg) was intrathecally administrated at 30 min before (pretreatment) or after (posttreatment) intrathecal injection of ephrinB1-Fc (0.5 μg), and then, we assayed the phosphorylation of three MAPKs for each treatment in mice spinal cord at 1 h after pretreatment and 30 min after posttreatment. We found that intrathecal injection of any one of the three MAPKs inhibitors only decreased the phosphorylation of its target kinase but not of the other two MAPKs kinases in naïve mice (fig. 6A). However, pretreatment with p38 inhibitor decreased the phosphorylation not only of p38 but also of ERK and JNK by intrathecal ephrinB1-Fc. The similar results were also found in pretreatment with ERK and JNK inhibitors (fig. 6B). Furthermore, posttreatment with JNK inhibitor decreased the phosphorylation of three MAPKs induced by intrathecal injection of ephrinB1-Fc, and posttreatment with p38 inhibitor decreased the phosphorylation of p38 and ERK, not JNK, and posttreatment with ERK inhibitor only decreased the phosphorylation of ERK (fig. 6C). These results suggested that decreased activation of two other MAPKs by any one of the MAPKs inhibitors resulted from inhibition of interaction of intracellular signaling pathways, not from direct inhibition of other MAPKs by this MAPKs inhibitor.

Spinal ephrinBs or EphBs signaling played a critical role in the development and maintenance of neuropathic and inflammation pain. To test whether spinal MAPKs activation contribute to the role of ephrinBs or EphBs signaling in neuropathic and inflammation pain, we analyzed the effect of intrathecal injection of EphB1-Fc, which can block ephrinBs or EphBs pathway, on expression of spinal p-MAPKs in CCI and formalin models (fig. 7). Consecutive intrathecal injection of EphB1-Fc on 1 h before CCI and days 1 and 2 after CCI (0.5 μg for each injection; see experimental protocol in fig. 7A) prevented the induction of thermal hyperalgesia and mechanical allodynia and this effect continued at least for 7 days after CCI (P < 0.05 or P < 0.01, EphB1-Fc + CCI vs.  Sal + CCI; fig. 7C). On day 5 after CCI, a peak time point of pain behaviors, L^4–5spinal cord of some mice were extracted for Fos protein and p-MAPKs expression. Injection of EphB1-Fc significantly inhibited the increased expression of spinal Fos protein (P < 0.01 vs.  Sal + CCI; fig. 7D) and p-MAPKs expression (Sal + CCI vs.  Sal + Sham: P < 0.001 for p-p38 and p-ERK, P < 0.05 for p-JNK; EphB1-Fc + CCI vs.  Sal + CCI: P < 0.05 for p-p38, p-ERK and p-JNK; fig. 7E) induced by CCI. Intrathecal injection of a single dose of EphB1-Fc (0.5 μg/5 μl) on fifth day after CCI (see experimental protocol in fig. 7B) reversed the established thermal hyperalgesia and mechanical allodynia, and this effect could last up to 8 h (P < 0.05, CCI + EphB1-Fc vs.  CCI + Sal; fig. 7F). Spinal Fos protein (P < 0.01 vs.  CCI + Sal; fig. 7G) and p-MAPKs (CCI + Sal vs.  Sham + Sal: P < 0.05 for p-p38, p-ERK and p-JNK; CCI + EphB1-Fc vs.  CCI + Sal: P < 0.05 for p-p38, p-ERK and p-JNK; fig. 7H) expression detected on 2 h after injection of EphB1-Fc also were markedly inhibited. We also got the similar antinociceptive effect when EphB1-Fc was injected in formalin model. Pretreatment with intrathecal injection of EphB1-Fc (0.5 μg/5 μl) at 30 min before formalin injection (see experimental protocol in fig. 8A) significantly reduced licking hind paw time in phase I and II responses (EphB1-Fc + Formalin vs.  Sal + formalin, P < 0.01 at phase II and P < 0.05 at phase I; fig. 8B). The spinal Fos protein (P < 0.01 vs.  Sal + formalin; fig. 8C) and p-MAPKs (Sal + formalin vs.  Sal + Sal: P < 0.05 for p-p38, p-ERK, and p-JNK; EphB1-Fc + formalin vs.  Sal + formalin: P < 0.05 for p-p38, p-ERK, and p-JNK; fig. 8D) expression, detected at 2 h and 30 min after injection of formalin, respectively, also was markedly inhibited.

NMDA receptor contributes to regulation of MAPKs activity in the spinal cord and many brain areas.43,44Activation of EphBs receptors could recruit and activate Src-family kinases, which then phosphorylate subunits of NMDA receptor and enhance activity of NMDA receptor.45,46Intrathecal administration of ephrinB2-Fc in adult rats, which can induce behavioral thermal hyperalgesia, led to NR2B tyrosine phosphorylation via  Src family kinases.14Moreover, a great deal of evidence has confirmed that NMDA receptor was involved in central sensitization induced by nociceptive stimuli.47–49Thus, we speculated that NMDA receptor may be involved in activation of MAPKs by intrathecal injection of ephrinB1-Fc. To test this hypothesis, MK-801 (2.5 μg), an NMDA receptor antagonist, was injected 30 min before administration of ephrinB1-Fc. The L^4–5of spinal cord was collected to detect the expression of Fos protein and p-MAPKs at 2 h and 30 min, respectively, after injection of ephrinB1-Fc (see experimental protocol in fig. 9A). Our data indicated that pretreatment with MK-801 significantly prevented thermal hyperalgesia and mechanical allodynia (MK-801 + ephrinB1-Fc vs.  ephrinB1-Fc, P < 0.05 from 0.5 to 4 h; fig. 9B) and spinal Fos protein expression (P < 0.01 vs.  ephrinB1-Fc; fig. 9C) and activation of spinal MAPKs (P < 0.05 vs.  ephrinB1-Fc; fig. 9D) induced by ephrinB1-Fc.

This study showed the following findings (1) spinal administration of ephrinB1-Fc produced a dose-dependent thermal hyperalgesia and mechanical allodynia, which was accompanied with the increased expression of spinal Fos protein and p-MAPKs; activated p38 and JNK were localized in both spinal neurons and astrocytes and p-ERK only in spinal neurons; (2) inhibition of spinal MAPKs prevented or reversed ephrinB1-Fc-induced hyperalgesia and spinal Fos protein expression; (3) blocking EphBs receptors alleviated formalin- and CCI-induced pain behaviors and inhibited activation of spinal MAPKs; and (4) spinal NMDA receptor mediated activation of spinal MAPKs by intrathecal injection of ephrinB1-Fc. These findings demonstrated a novel mechanism for ephrinBs/EphBs system involved in spinal nociceptive information modulation.

Central sensitization, an activity-dependent functional plasticity in spinal cord neurons, is one of the main causes of behavior hyperalgesia under pathologic conditions and has been under intensive investigation.50–52Activation of postsynaptic membrane receptors or ion channels, intracellular kinase cascades, and intranuclear gene expression contributes to the induction, development, and maintenance of central sensitization. Several studies have suggested that central sensitization related to pain and hippocampal long-term potentiation (LTP) associated with learning and memory may share certain mechanisms.52,53For example, activation of NMDA receptor and the subsequent associated intracellular signal transduction cascades are involved in the induction, development, and maintenance of synaptic plasticity in the spinal cord and hippocampus.

Both in vitro  and in vivo  studies had provided evidence that ephrinBs/EphBs system participated in synaptic plasticity and learning and memory in the adult hippocampus. EphB2 knockout mice show deficits in NMDA-dependent synaptic plasticity in the hippocampus, such as LTP and long-term depression.46,54Furthermore, mossy fiber LTP is also impaired by extracellular application of ephrinB1-Fc to inhibit EphB interactions with B-ephrins that are presumably localized presynaptically.55Now that there are some common cellular and molecular mechanisms between LTP in the hippocampus and central sensitization of nociceptive neurons in the dorsal horn, whether both share the ephrinBs/EphBs system mechanism is unknown. Our previous and current studies addressed this question. Application of EphB1-Fc, which can block EphBs receptors, onto the spinal cord significantly reduced the enhanced responses of wide dynamic range neurons in the dorsal horn to the innocuous and noxious stimulation in CCI rats by means of extracellular recordings in rat dorsal horn in vivo .17C-fibers-evoked LTP of the dorsal horn neurons may be associated with inflammation- and nerve injury-induced process of central sensitization and behavioral pain syndromes. We found that activation of ephrinBs-EphBs receptor signaling is required for induction of c-fibers-evoked LTP of dorsal horn neurons and lowers stimuli threshold of the LTP.17 

Fos protein expression has been used as a maker for neuronal activation in the central nervous system and has been widely used as a tool in functional mapping of neuronal circuits in response to various defined stimuli. There is a positive correlation between the quantity of Fos protein expression and the degree of sensitization induced by nociceptive stimuli in spinal cord neurons. Long-term facilitation of c-fiber-evoked firing of wide dynamic range neurons in the spinal dorsal horn in response to conditioning stimulation of afferent fibers was accompanied with the frequency-dependent increase of c-fos-labeled cells in superficial, intermediate, and deep laminae of the dorsal horn on the stimulated side.56Inhibition of tetanically sciatic stimulation-induced LTP of spinal neurons was also accompanied with the suppressed tetanic stimulation-induced spinal Fos expression.57In this study, we further confirmed the previous electrophysiologic findings by means of Fos protein expression. Spinal application of ephrinB1-Fc, which can activate EphBs receptors, induced a dose-dependent increase in spinal Fos protein expression. Blocking of EphBs receptors with intrathecal injection of EphB1-Fc also suppressed the increased expression of spinal Fos protein expression in CCI and formalin pain models. Although we cannot identify which EphBs subgroup was activated dominantly by ephrinB1-Fc, our study demonstrates clearly that activation of ephrinBs/EphBs system is required for development of the pain-related, long-lasting alterations of excitability and synaptic plasticity of spinal neurons, which lead to increased sensitivity to noxious stimuli (hyperalgesia) or nonnoxious stimuli (allodynia).

EphBs receptors transduce signals bidirectionally by interacting with membrane-anchored ephrinBs ligands expressed on adjacent cells. Application of ephrinB1-Fc reagent can activate EphBs receptors (forward signaling) and also prevent activation of endogenous ephrinBs ligands (reverse signaling). EphB1-Fc can disturb the signaling by the interaction of EphBs and ephrinBs and activate endogenous ephrinBs ligands. In this study, we could not confirm whether reverse signaling produced by ephrinBs activation was involved in hyperalgesia induced by activation of spinal ephrinBs/EphBs signaling. However, injection of EphB1-Fc could not induce any pain behavior and not affect basal pain threshold. It is possible that reverse signaling has no contribution to this process, or EphB1-Fc induces only weak agonist activity.

A great deal of evidence has indicated that spinal glial activation contributes to initiation and maintenance of central sensitization.58It has been widely demonstrated that expression of Eph receptors or their ephrin in astrocytes played multiple pivotal roles in various physiologic and pathologic processes such as hippocampus plasticity and spinal cord injury.59–61Ephrin/Eph signaling has been shown to be a potential regulator of neuron–astrocyte interactions mediating rapid structural and functional plasticity.62,63This study showed that intrathecal injection of ephrinB1-Fc induced activation of p38 and JNK in spinal astrocytes, indicating that ephrinBs/EphBs signaling in spinal astrocytes was involved in nociceptive information modulation in this pain model. However, further researches are required to address the mechanisms of spinal astrocytes involved in development and maintenance of pain behavior related to ephrinBs/EphBs signaling.

Activation of intracellular kinase cascades and intranuclear gene expression plays a critical role in the development and maintenance of central sensitization. According to the structural features of EphBs receptors, MAPKs signaling may be a good candidate as its downstream effectors. As other receptor tyrosine kinases, EphBs receptors contain a highly conserved motif containing two tyrosine residues that are the major autophosphorylation sites. Activation of EphBs receptors induces autophosphorylation of specific tyrosine residues and creates docking sites for downstream molecules that contain Src homology 2 domain, such as Shc adaptor protein, and then Shc bridges EphBs to Grb2, which leads to activation of several downstream intracellular signaling, including the MAPKs cascade. On the other hand, activated EphBs receptors can also modulate the NMDA receptor function via  a Src-dependent mechanism and then enhance NMDA receptor-mediated Ca²⁺. This [Ca²⁺]i increase in turn directly facilitates the MAPKs pathways.

MAPKs, including p38, ERK, and JNK, are a family of serine/threonine protein kinases that transduce extracellular stimuli into intracellular posttranslational and transcriptional responses. It is well established that the MAPKs activation may be involved in the modulation of nociceptive information and peripheral and central sensitization produced by intense noxious stimuli through various routes.18–23Some studies report that ephrinBs/EphBs signaling regulates the activities of small GTPases of the Ras family and then activates the Ras/MAP kinase pathway in a wide variety of cell lines and tissues.24–27,37,38For example, ephrinB2 stimulation of EphB receptors in T cells also activates ERK1 and ERK2 MAP kinases. Ectopic expression of ephrinB1 in 293T cells leads to JNK activation.37EphB1 recruits c-Src and p52 Shc to activate MAPK/ERK and promotes chemotaxis.64These studies suggested that MAPKs signaling may mediate the role of ephrinBs/EphBs signaling in some physiologic and pathologic conditions. This study further supported and extended this view. We found that intrathecal injection of ephrinB1-Fc induced hyperalgesia accompanied by activation of spinal MAPKs. Blocking MAPKs prevents and reversed pain behavior and spinal Fos expression induced by intrathecal injection of ephrinB1-Fc. Furthermore, inhibiting spinal EphB1 receptor by intrathecal injection of ephB1-Fc reduces neuropathic and inflammation pain accompanied by the decreased expression of spinal MAPKs and Fos protein. These findings indicated that MAPK pathway was involved in spinal nociceptive information modulation related to ephrinBs/EphBs signaling. We further showed that pretreatment with NMDA receptor antagonist MK-801 prevented the activation of spinal MAPKs by ephrinB1-Fc accompanied by the reduced pain behavior. It suggested that NMDA receptor contributed to the activation of spinal MAPKs by ephrinB1-Fc. The activated MAPKs can translocate from cytoplasm into nucleus and, in turn, phosphorylate and increase the expression of transcriptional factors, such as cyclic adenosine 3′,5′-monophosphate response element binding and Fos and promotes the transcription of downstream genes. This point could reasonably interpret why ephrinB1-Fc-induced hyperalgesia can last at least up to 24 h, when activation of MAPKs almost backed to baseline level at 8 h after intrathecal injection of ephrinB1-Fc.

This study found that intrathecal injection of ephrinB1-Fc activated all three MAPKs pathways; however, inhibition of any one of the three MAPKs pathways could have almost completely prevented or reversed pain behavior and Fos expression induced by intrathecal injection of ephrinB1-Fc. How to interpret this result? We believed that nociceptive stimuli activated a network of signaling molecules in the spinal cord. Each signaling molecule interacts with others in this network. Depending on the interaction of signaling molecules, this network keeps a high-level activation status, which could initiate and maintain pain behavior. Blocking any one of the interactions in this network will result in the decreased activation of whole network and pain behavior relief as a consequence. This viewpoint was supported by the results of the field of pain study. For example, lots of signaling pathways and molecules were involved in the development and maintenance of various types of pain. However, pain behaviors can almost completely be prevented or reversed by blocking any one of these signaling pathways or molecules. In agreement with this point, our study showed that inhibition of any one of the three MAPKs pathways also decreased the activation of the other two MAPKs.

This study demonstrates that activation of spinal ephrinBs/EphBs system induces hyperalgesia through an MAPKs-mediated mechanism. These findings may have important implications for exploring the roles and mechanisms of ephrinBs/EphBs system underlying central nervous system disease and for understanding the molecular basis for underlying physiologic and pathologic pain. In addition, it suggests that ephrinBs/EphBs pathway would be a new potential target for treatment of pathologic pain.