The reduction of γ-aminobutyric acid (GABA) type A receptor–mediated inhibition has long been implicated in spinal sensitization of nociceptive responses. However, it is largely unknown which signaling cascades in spinal dorsal horn neurons are initiated by the reduced inhibition to trigger pain hypersensitivity.
GABAergic inhibition was manipulated by intrathecal application of GABA type A receptor antagonist bicuculline in intact mice or by GABA type A receptor agonist muscimol in complete Freund’s adjuvant–injected mice. Immunoblotting, coimmunoprecipitation, immunohistochemistry, and behavioral tests were used to explore the signaling pathways downstream of the altered GABAergic tone.
The study data revealed that the 61-kD isoform of striatal-enriched protein phosphatase (STEP61) was a key molecule that relayed the signals from GABAergic neurotransmission. The authors found that STEP61 was highly expressed in dorsal horn neurons. Under physiological conditions, STEP61 tonically interacted with and negatively controlled the activities of extracellular signal–regulated kinase and Src-family protein tyrosine kinases member Fyn, two critical kinases involved in spinal sensitization. Once GABAergic inhibition was impaired, STEP61 interaction with its substrates was substantially disturbed, allowing for activation of extracellular signal–regulated kinase and Fyn (n = 4 to 6). The hyperactivities of extracellular signal–regulated kinase and Fyn, along with STEP61 dysregulation, caused the tyrosine phosphorylation and synaptic accumulation of GluN2B subunit-containing N-methyl-d-aspartate subtype of glutamate receptors (n = 6), leading to GluN2B receptor-dependent pain hypersensitivity. Overexpression of wild-type STEP61 to resume its enzymatic activity significantly blocked the mechanical allodynia evoked by bicuculline and more importantly, alleviated chronic inflammatory pain (n = 6 in each group).
These data identified STEP61 as a key intermediary for GABAergic inhibition to regulate pain sensitization.
STEP61 is expressed in spinal cord dorsal horn neurons. γ-Aminobutyric acid type A receptors work through STEP61 to regulate extracellular signal–regulated kinase and Src-family protein tyrosine kinases member Fyn. STEP61 dysfunction augments glutamate receptor function.
γ-Aminobutyric acid type A receptors are involved in nociceptive processing within spinal cord dorsal horn
61-kD isoform of striatal-enriched protein phosphatase (STEP61) inhibits several signaling cascades linked to nociceptive sensitization and chronic pain
61-kD isoform of striatal-enriched protein phosphatase (STEP61) is expressed in spinal cord dorsal horn neurons
γ-Aminobutyric acid type A receptors work through STEP61 to regulate extracellular signal–regulated kinase and Src-family protein tyrosine kinases member Fyn
STEP61 dysfunction augments glutamate receptor function
γ-AMINOBUTYRIC acid (GABA) type A receptor (GABAAR) within spinal dorsal horn mediates the fast inhibitory synaptic transmission, which is engaged in spinal processing of nociceptive information. After tissue and nerve injury, GABAAR-mediated inhibition is rapidly and persistently reduced as a result of decreased presynaptic GABA release and/or disturbed postsynaptic anion homeostasis.1,2 The reduced inhibition is widely considered as a critical player in spinal sensitization of painful behaviors.3,4 Reinforcing GABAergic neurotransmission potently ameliorates pathological pain symptoms, whereas direct blockade of GABAAR in intact animals mimics the peripheral injuries by reducing the pain thresholds in response to innocuous stimuli.3,5,6 Nevertheless, the signaling cascades that are initiated by the reduced inhibition to trigger the pain hypersensitivity remain to be characterized.
Striatal-enriched protein phosphatase (STEP) is a brain-specific tyrosine phosphatase that plays an important role in synaptic plasticity and several neuropsychiatric disorders.7 STEP physically associates with and inactivates extracellular signal–regulated kinase 1/2 (ERK1/2), p38 mitogen-activated protein kinase (p38-MAPK), and Src-family protein tyrosine kinases (SFKs) member Fyn,7 three critical molecules involved in spinal sensitization. Direct inhibition of ERK1/2, Fyn, or p38-MAPK produces antihyperalgesic or antiallodynic effects in several animal models of pain.8–11 In addition to these protein kinases, STEP also interacts with N-methyl-d-aspartate subtype of glutamate receptors (NMDARs), especially those containing GluN2B subunits (GluN2B receptors).7 Accumulating evidence has indicated that peripheral lesions can specifically recruit GluN2B receptors at excitatory glutamatergic synapses, leading to the potentiation of GluN2B receptor–mediated nociceptive transmission.12 Through dephosphorylating GluN2B at Tyr1472, STEP has been shown to internalize GluN2B receptors from plasma membrane, depress NMDAR-mediated synaptic currents, and block the induction of NMDAR-dependent long-term potentiation, one of the best-characterized forms of neuronal plasticity pertaining to spinal sensitization.7,13 Of the most importance is that the inhibition conferred by STEP on its substrates is dynamically regulated by a broad range of extracellular stimuli.14,15 For example, STEP phosphorylation by cyclic adenosine 3’,5’-monophosphate–dependent protein kinase (PKA) can reduce its enzyme activity and disturb its interaction with substrates.7,14 The altered STEP function has widespread implications for learning, memory, and neuropsychiatric disorders.7 The current study found that 61-kD isoform of STEP (STEP61) was the unique STEP variant that was present in spinal dorsal horn neurons. Through investigating the influence of the reduced GABAergic inhibition on spinal STEP61 function, we tested the hypothesis that STEP61 might serve as a key intermediary for GABAergic inhibition to regulate the pain hypersensitivity.
Materials and Methods
Animals and Drugs
All experimental procedures were in accordance with the guidelines of the Animal Care and Use Committee of Lanzhou University. Male adult C57BL/6J mice (18 to 22 g), provided by the Experimental Animal Center of Lanzhou University, were housed three to four per cage with free access to food and water. The animals were randomly assigned to different groups. Complete Freund’s adjuvant (CFA; 10 μl; Sigma, St. Louis, MO) was injected into the plantar surfaces of hind paws to induce inflammatory pain. For intrathecal drug delivery,16 the mice were held firmly by a pelvic girdle and a 30-gauge needle attached to a 25-μl microsyringe was inserted between L5 to L6 vertebrae. A sudden advancement of the needle accompanied by a slight flick of the tail was used as the indicator for the proper insertion into the subarachnoid space. Bicuculline, muscimol, ifenprodil, H-89, U-0126, SB203580 (Sigma), or PP2 (Calbiochem, La Jolla, CA) was intrathecally injected slowly in 5-μl volume. The recombinant adenovirus (1010 pfu/ml) expressing green fluorescent protein (GFP)–tagged STEP61 or its catalytically inactive STEP61(C472S) mutant (cysteine to serine mutation at residue 472) was commercially obtained from Yingrun Biotechnologies (Changsha, China) and intrathecally given in 5-μl volume.17 All the experiments were conducted blindly to the experimenters without knowledge of the manipulations that the animals received.
Mice were habituated in a cage with wire mesh floor for at least 30 min before pain sensitivity was measured. A set of eight Von Frey filaments (Stoelting, Wood Dale, IL) was applied to the plantar surfaces of hind paws and 50% paw withdrawal threshold (PWT) was calculated by using up–down method as described previously.5
Subcellular Fractionation, Immunoprecipitation, and Western Blot
The mice were anesthetized with sodium pentobarbital (60 mg/kg, intraperitoneally) and the lumbar enlargements of spinal cords were quickly removed into ice-cold artificial cerebrospinal fluid (119.0 mM NaCl, 2.5 mM KCl, 2.5 mM CaCl2, 1.3 mM MgCl2, 1.0 mM NaH2PO4, 26.0 mM NaHCO3, 11.0 mM d-glucose, pH 7.4, bubbled with 95% O2 + 5% CO2). The dorsal quadrants of spinal cords were dissected out and homogenized in the lysis buffer (10.0 mM Tris·HCl, pH 7.6, 320.0 mM sucrose, 5.0 mM EDTA, and proteases/phosphatases inhibitors [10.0 mM NaF, 1.0 mM orthovanadate, 1.0 mM phenylmethylsulfonyl fluoride, and 1.0 mg/ml each of aprotinin, chymostatin, leupeptin, antipain, and pepstatin]). The homogenates were centrifuged at 1,000g for 10 min at 4°C to remove the nuclei and large debris (P1). The supernatant (S1) was collected and centrifuged at 10,000g for 15 min to obtain the P2 pellet that contained the crude synaptosomal fraction. P2 was incubated for 30 min with the lysis buffer containing 0.5% Triton X-100 and then centrifuged at 32,000g for 20 min to harvest the synaptosomal membrane fraction (P3).18,19 To assay the protein expression and phosphorylation,10 the spinal dorsal horn was homogenized in radio-immunoprecipitation assay buffer (50.0 mM Tris·HCl, pH 8.0, 150.0 mM NaCl, 1.0 mM EDTA, 1.0% Nonidet P-40, 0.1% sodium dodecyl sulfate, 0.5% sodium deoxycholate, and phosphatases/proteases inhibitors). After centrifugation at 14,000g for 10 min, the supernatant was collected and the protein concentration was measured by using bicinchoninic acid assay kit (Pierce, Rockford, IL).
For coimmunoprecipitation,20 P2 fraction was extracted in 50.0 mM Tris·HCl, pH 9.0, 10.0 mM EDTA, 1.0% sodium deoxycholate, and proteases/phosphatases inhibitors at 37°C for 30 min. Equal volume of the dilution buffer (50.0 mM Tris·HCl, pH 7.4, 150.0 mM NaCl, 0.1% sodium dodecyl sulfate, 1.0% Triton X-100, and proteases/phosphatases inhibitors) was added into the extract above. After centrifugation at 14,000g, the supernatant was collected and incubated at 4°C with anti-STEP antibody overnight. The protein A/G-agarose beads were incubated with the immune complexes for 4 h. After three washes, the immunoprecipitates were resuspended in sodium dodecyl sulfate sample buffer and boiled for 5 min before Western blot analysis. For immunoprecipitation of Fyn and Src,21 P2 fraction was lysed in radio-immunoprecipitation assay buffer, followed by centrifugation at 14, 000g for 10 min. The supernatant was incubated with anti-Fyn or anti-Src antibody overnight at 4°C. The immune complexes were precipitated by protein A/G-agarose beads.
The equal amounts of protein samples were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. After blocking with 5% nonfat milk, the membranes were incubated with individual primary antibody overnight at 4°C, followed by incubation with horseradish peroxidase–conjugated secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA). The blots were visualized by enhanced chemiluminescence (Beyotime Institute of Biotechnology, Jiangsu, China). The primary antibodies used in the current study included the rabbit anti-GluN2B, rabbit anti-GluN2A, rabbit anti-GluN2B-pY1472, and mouse anti-Src antibody from Millipore (Temecula, CA); mouse anti-GluN1 antibody from BD Pharmingen (San Diego, CA); mouse anti-β-actin antibody from Sigma; goat anti-Fyn antibody from Santa Cruz (Santa Cruz, CA); rabbit anti-Src-pY418 antibody from Invitrogen (Camarillo, CA); mouse anti-STEP antibody from Upstate (Lake Placid, NY); and rabbit anti-ERK1/2, mouse anti-ERK1/2-pThr202/Tyr204, rabbit anti-p38-MAPK-pThr180/Tyr182, and rabbit anti-p38-MAPK antibody from Cell Signaling (Beverly, MA).
Mice were perfused through the ascending aorta with phosphate-buffered saline (0.01 M) followed by 4% paraformaldehyde in 0.1 M phosphate buffer. The L4 to L5 spinal cord segments were dissected and fixed in the same fixative for 4 h. The tissues were cryoprotected in 30% sucrose overnight. Transverse sections (16 μm) were cut and incubated with mouse anti-STEP antibody and rabbit anti-NeuN antibody (Anbo Biotechnology, JiangSu, China), rabbit anti-Iba1 antibody (Wako, Osaka, Japan), or rabbit antiglial fibrillary acidic protein antibody (Dako, Glostrup, Denmark) at 4°C for 72 h. The sections from mice that received intrathecal viral injection were incubated with anti-NeuN antibody. After several washes, the sections were incubated with Alexa Fluor 488-conjugated and/or Cy3-conjugated secondary antibodies for 2 h before images were captured.22
All data were represented as mean ± SEM. Differences between groups were compared by using Student t test and one-way ANOVA followed by post hoc Tukey honestly significant difference test. If an experiment design involved more than one independent variable (intradermal injection, intrathecal injection, or time), the statistically significant interaction between the effects of independent variables on the dependent variable was first determined by two-way or three-way ANOVA, followed by one-way ANOVA and post hoc Tukey honestly significant difference test to compare the differences between groups. The criterion for statistical significance was P value less than 0.05.
Role of STEP61 in the Pain Hypersensitivity Evoked by GABAAR Antagonist Bicuculline
The STEP family consists of several members, including the transmembrane 61-kD isoform (STEP61) and the cytosolic 46 and 38 kD isoforms.7 In the spinal cord of adult mice, the most abundant STEP variant was STEP61 (fig. 1A). To define the cellular localization of STEP61 in spinal dorsal horn, we performed double immunofluorescence with antibodies against STEP and neuronal marker NeuN, astrocyte marker glial fibrillary acidic protein, or microglia marker Iba1. The images illustrated that most of STEP61 coincided with NeuN, but not with glial fibrillary acidic protein and Iba1 (fig. 1B), suggesting that STEP61 was predominantly expressed in spinal neurons. To investigate the possible role of STEP61 in spinal processing of nociceptive signals, we manipulated endogenous STEP61 activity by intrathecally injecting recombinant adenovirus that encoded GFP-tagged wild-type STEP61 (STEP61(WT)) or its dominant-negative STEP61(C472S) mutant. The recombinant adenovirus successfully infected dorsal horn neurons, as evidenced by GFP signals that coincided with NeuN (fig. 1C). Immunoblotting analysis revealed that the recombinant adenovirus encoding STEP61(WT) and STEP61(C472S) increased the protein levels of STEP61 to 313.2 ± 12.5% of control (P < 0.05, n = 6 experiments) and 306.6 ± 7.3% of control (P < 0.05, n = 6 experiments), respectively (fig. 1D). The behavioral tests showed that viral expression of GFP or STEP61(WT) produced no effects on the PWT values of intact mice (fig. 1E). However, spinal expression of catalytically inactive STEP61(C472S) elicited a pronounced allodynia in intact mice (fig. 1E). The reduction of PWT values was observed 1 day after viral injection, reached maximum within 4 days, and lasted for 12 days (fig. 1E). This long-lasting decline in pain thresholds observed with STEP61(C472S), rather than STEP61(WT), implicated that endogenous STEP61 might exert a persistent and sufficient inhibition of nociceptive responses at physiological conditions.
The traditional gate control theory of pain has conferred on GABAergic neurotransmission a permissive role in the sensation of allodynia.6 Brief removal of spinal inhibition by selective GABAAR antagonist bicuculline mimics many symptoms of pathological pain.3–5 The current study showed that intrathecal bicuculline (0.1 μg) application significantly decreased PWT values of GFP-expressing mice (fig. 1F). However, prior expression of STEP61(WT) totally blocked the pronociceptive action of bicuculline (fig. 1F), suggesting that the pain hypersensitivity caused by the reduced inhibition might involve STEP61 dysfunction.
GABAAR Blockade Disrupted STEP61 Interaction with ERK1/2 and Fyn
Previous studies have indicated that STEP61 function can be attenuated through two pathways: the proteolytic cleavage or ubiquitination that decreases STEP61 protein level and the phosphorylation by PKA that not only inhibits STEP61 activity but also reduces its binding affinity for substrates.7 To explore the molecular mechanisms for the reduced GABAergic neurotransmission to perturb STEP61 function, we first examined the possible change of STEP61 protein expression, finding that bicuculline generated no significant influence on total STEP61 protein level (fig. 2A). It seemed that STEP61 dysfunction was not attributed to its protein degradation. Because STEP61 binds and inhibits ERK1/2, p38-MAPK, and Fyn that mediate the effects of STEP61 on a wide range of cellular responses,7 we next assayed the potential influence of GABAAR blockade on STEP61 complex formation. In saline-injected control mice, anti-STEP antibody precipitated ERK1/2, Fyn, and p38-MAPK from crude synaptosomal fraction of spinal dorsal horn (fig. 2B), suggesting that STEP61 interacted with these signaling components at resting conditions. Acute bicuculline treatment for 40 min, however, dramatically reduced Fyn and ERK1/2 amounts in STEP precipitates (fig. 2B). This molecular dissociation was specific because p38-MAPK contents pulled down by anti-STEP antibody displayed no detectable difference between saline-injected and bicuculline-injected mice (fig. 2B). Meanwhile, this dissociation was reversible and required the activation of PKA.14 When PKA inhibitor H-89 (2.5 μg) was intrathecally superimposed at 10-min postbicuculline injection, it simultaneously alleviated bicuculline-induced allodynia (data not shown) and resumed STEP61 interaction with ERK1/2 and Fyn (fig. 2B). In the absence of bicuculline, H-89 alone had no effect on STEP61 complex (fig. 2C). Given that the physical interaction with Fyn and ERK1/2 is a critical step for STEP61 to dephosphorylate and negatively control these pain-related kinases, the disruption of STEP61 complex by the reduced GABAergic neurotransmission might cause STEP61 dysfunction in spinal dorsal horn.
GABAAR Blockade Initiated ERK and Fyn Signalings in Spinal Dorsal Horn via STEP61 Dysfunction
Our data demonstrated that the dissociation of STEP61 complex by bicuculline removed STEP61-mediated inhibition and allowed the activation of ERK and Fyn. As shown in figure 3, A and B, a significant increase in the phosphorylation of ERK2 at Thr185/Tyr187 and of Fyn at Tyr420 was elicited by bicuculline. ERK1 showed a tendency to increase its phosphorylation after bicuculline application, which, however, failed to reach statistical significance (fig. 3A). Spinal overexpression of exogenous STEP61(WT), which was predicted to interact with endogenous ERK/Fyn and reinforce STEP61-mediated inhibition, substantially attenuated the stimulatory effects of bicuculline on ERK2 and Fyn (figs. 3, A and B). By comparison, the phosphorylation of Src at Tyr418 (fig. 3C) and of p38-MAPK at Thr180/Tyr182 (fig. 3D) exhibited no significant changes after either bicuculline or STEP61(WT) treatment. STEP61(WT) alone did not affect basal ERK1/2 and Fyn phosphorylation in intact mice (data not shown). Behavioral tests showed that MAPK inhibitor U-0126 (fig. 4A) and SFKs inhibitor PP2 (fig. 4B), when intrathecally given at 10-min postbicuculline injection, dose dependently attenuated the reduction of PWT values, whereas p38-MAPK inhibitor SB203580 had no effect (fig. 4C). These data implicated that the reduced GABAergic inhibition, through dissociating STEP61 complex, stimulated ERK and Fyn activities, causing Fyn/ERK-dependent pain hypersensitivity.
GABAAR Blockade Enhanced the Function of NMDAR via STEP61 Dysregulation
The reduced GABAergic neurotransmission has been shown to promote the synaptic accumulation of GluN2B subunit-containing NMDAR, a critical player in nociceptive plasticity.5 Intrathecal application of GluN2B-selective antagonist ifenprodil dose dependently ameliorated the mechanical allodynia induced by bicuculline (fig. 5A). Given that STEP61 negatively regulates GluN2B tyrosine phosphorylation and synaptic clustering,7 we tested whether bicuculline-induced GluN2B receptor hyperfunction resulted from STEP61 dysregulation. In GFP-expressing mice, brief removal of GABAergic inhibition elicited a robust phosphorylation of GluN2B at Tyr1472 (GluN2B-Y1472) (fig. 5B), coincident with which were the increased immunoreactivities of GluN2B and GluN1 subunits at synaptosomal membrane fractions (fig. 5C). Expression of STEP61(WT), however, prevented GluN2B-Y1472 phosphorylation (fig. 5B) and blocked GluN2B/GluN1 synaptic accumulation induced by bicuculline (fig. 5C). The synaptic content of NMDAR GluN2A subunit was insensitive to bicuculline or STEP61(WT) treatment (fig. 5C). It was noteworthy that, besides STEP61 per se, the active STEP61 substrates Fyn and ERK1/2 might also contribute to bicuculline-evoked GluN2B hyperfunction. Intrathecal administration of SFKs inhibitor PP2 or MAPK inhibitor U-0126 produced a similar inhibition of GluN2B synaptic expression in bicuculline-injected mice as STEP61(WT) (fig. 5D), whereas p38-MAPK inhibitor SB203580 had no effect (fig. 5D). These data suggested that STEP61 dysregulation, together with Fyn/ERK activation, led to GluN2B hyperfunction after GABAAR blockade.
Peripheral Inflammation Disrupted STEP61 Complex via the Reduced GABAergic Inhibition
Peripheral tissue injury naturally impairs spinal GABAergic inhibition, which is considered as a key initiator of chronic inflammatory pain.3–5 However, the means by which the reduced inhibition aggravates pathological pain are unclear. To test whether the reduced inhibition disturbed STEP61 signaling during inflammatory pain, we performed coimmunoprecipitation with anti-STEP antibody 1 day after intraplantar CFA injection. Compared with saline control, CFA-injected mice displayed much less amounts of Fyn and ERK1/2 in STEP precipitates (fig. 6A). There was no detectable reduction of precipitated p38-MAPK content in inflamed mice relative to control littermates (fig. 6A). When a selective GABAAR agonist muscimol (0.1 μg) was intrathecally applied for 30 min to ameliorate inflammatory pain (data not shown),5 the interaction of STEP61 with ERK1/2 and Fyn was substantially resumed (fig. 6A). Similar to muscimol, spinal treatment with PKA inhibitor H-89 (2.5 μg) for 30 min also relieved inflammatory pain10 and restored STEP61 binding to Fyn and ERK1/2 (fig. 6A). CFA, muscimol, and H-89 did not affect the protein expression of STEP61 at homogenates (fig. 6B), suggesting that the disruption of STEP61 complex by the reduced inhibition correlated with inflammatory pain.
Previous studies have indicated that CFA induces a persistent increase of ERK1/2 phosphorylation in spinal dorsal horn neurons, which increases pain sensitivity through several mechanisms involving gene expression.23 Our data confirmed this by showing a significant increase of ERK1/2 phosphorylation 1 day after CFA injection in mice (fig. 7A). Importantly, spinal treatment with muscimol to resume STEP61/ERK1/2 binding repressed ERK1/2 phosphorylation of inflamed mice to control level (fig. 7A). Similarly, the enhancement of STEP61/Fyn binding by muscimol also inhibited CFA-induced Fyn phosphorylation (fig. 7B), suggesting that the reduced GABAergic tone served as an important stimulant of ERK1/2 and Fyn activities during inflammatory pain. Meanwhile, the tyrosine phosphorylation and synaptic accumulation of GluN2B in inflamed mice were also significantly suppressed by muscimol (fig. 7C). The p38-MAPK phosphorylation experienced no significant change after either CFA or muscimol injection (fig. 7D).
Expression of Wild-type STEP61 Ameliorated Inflammatory Pain
If the reduced GABAergic inhibition disturbed STEP61 signaling to exaggerate inflammatory pain, the enhancement of STEP61 activity might yield a similar analgesic action as GABAAR agonists.3–5 To test this, the recombinant adenovirus encoding STEP61(WT) was intrathecally administrated, followed by intraplantar CFA injection. Successive monitoring of nociceptive responses revealed that prior expression of STEP61(WT) for 3 days blunted the reduction of PWT values induced by CFA (fig. 8A). At each time point tested after CFA injection, the mechanical thresholds were significantly higher in STEP61(WT)-expressing mice than those in GFP-expressing ones (fig. 8A). Moreover, when the recombinant adenovirus encoding STEP61(WT) was intrathecally injected at 2 h post-CFA, it also alleviated the established allodynia. As shown in figure 8B, a significant increase of PWT values in inflamed mice was observed 1 day after viral injection, which lasted until the end of experiments.
The major finding in the current study was that a multifunctional protein tyrosine phosphatase STEP61 acted downstream of spinal GABAergic neurotransmission and gated the nociceptive responses to innocuous stimuli. We provided evidence that the maintenance of basal STEP61 activity required normal GABAergic neurotransmission. Once removing spinal inhibition by bicuculline or peripheral inflammation, the negative control by STEP61 over its substrates, ERK and Fyn, was dramatically reduced, resulting in ERK/Fyn activation. Our data demonstrated that STEP61 dysfunction, resulting from its dissociation with substrates, was an important step for the reduced GABAergic inhibition to trigger spinal sensitization. Activation of GABAAR to resume STEP61 complex formation or direct overexpression of STEP61(WT) greatly blocked the generation and development of inflammatory pain.
The nociceptive neurons in spinal dorsal horn have been shown to receive the synaptic inputs from glutamatergic primary sensory neurons and local GABAergic interneurons. The functional equilibrium between excitatory and inhibitory neurotransmission is important for the integration of sensory signals. Previous studies have indicated that NMDAR blockade can effectively alleviate mechanical allodynia caused by GABAAR inhibition,5 suggesting an intimate crosstalk between GABAAR and NMDAR. In support of this notion, our data showed that GABAAR antagonist bicuculline significantly enhanced the synaptic expression of GluN2B receptors. Conversely, GABAAR agonist muscimol ameliorated the inflammatory pain by repressing synaptic GluN2B contents. We revealed that STEP61, one of the important postsynaptic phosphatases that exert tonic inhibition of glutamatergic synaptic transmission,24 was essential for the signal propagation from GABAAR to NMDAR. By disturbing STEP61 function, the reduced GABAergic inhibition exacerbated the nociceptive conveyance mediated by NMDAR. Interestingly, NMDAR activity has also been reported to regulate GABAAR properties. In hippocampal neurons, Ca2+ influx via NMDAR disperses GABAAR out of synapses, which reduces the inhibitory synaptic strength with important implications for synaptic plasticity and information processing.25 NMDAR activation also initiates protein phosphatase-1 signaling to internalize the surface potassium chloride cotransporter 2, leading to the switch of hyperpolarizing GABAergic transmission to be depolarizing and excitatory.26 These findings, in combination with our results, raised the possibility that NMDAR and GABAAR counteracted each other in tuning the neural network activity. In spinal dorsal horn, a positive feedback loop between the reduced GABAergic inhibition and NMDAR hyperfunction might account for the persistent sensitization of painful behaviors after peripheral lesions.
In addition to NMDAR, SFKs member Fyn and ERK1/2 are also subjected to the negative control by STEP61.7 Our data showed that selective GABAAR blockade by bicuculline disrupted STEP61 interaction with Fyn and ERK1/2, a process that was attributed to PKA activation.14 Pharmacological inhibition of PKA activity resumed STEP61 binding to these kinases in bicuculline-injected mice. The mechanisms for the reduced inhibition to activate PKA might involve the depolarization of plasma membrane, which has been shown to drive Ca2+ influx via voltage-operated Ca2+ channel and trigger Ca2+-dependent intracellular responses such as PKA activation.5,27,28 STEP61 complex dissociation by PKA might attenuate STEP61-mediated inhibition and permit the activation of Fyn and ERK. We found that intrathecal bicuculline application significantly increased Fyn autophosphorylation. However, bicuculline seemed to preferentially enhance ERK2 phosphorylation in intact mice. Presumably, endogenous STEP61 exerted a weaker inhibition on ERK1 than ERK2 so that the disruption of STEP61 complex by bicuculline generated differential influence on ERK1 and ERK2 phosphorylation. Somewhat distinct from bicuculline, CFA enhanced both ERK2 and ERK1 phosphorylation, possibly because the mechanisms for peripheral lesions to reduce GABAergic inhibition were far more complex than brief GABAAR blockade.3,4,29 Nevertheless, ERK1 phosphorylation is dispensable for spinal sensitization.9
As the STEP61 substrates, NMDAR, Fyn, and ERK2, have all been implicated in spinal nociceptive plasticity. Inhibition of NMDAR, especially GluN2B receptor, blocks long-term potentiation of glutamatergic nociceptive transmission, an important feature of spinal sensitization, and attenuates pathological pain.30–33 SFKs inhibition also reduces the NMDAR component of synaptic transmission, blocks NMDAR-dependent long-term potentiation, and alleviates pain hypersensitivity.10,21,34 With respect to ERK1/2, there have been some studies showing that ERK2 plays a more dominant role than ERK1 in long-term potentiation process.35 Specific ERK2 knockdown in spinal dorsal horn blocks the nociceptive plasticity induced by peripheral tissue injury.8 Our data demonstrated that GluN2B, Fyn, and ERK2 hyperactivities, resulting from STEP61 dysfunction, were critical for the reduced GABAergic inhibition to cause pain sensitization. Inhibition of GluN2B, Fyn, or ERK effectively attenuated the mechanical allodynia in bicuculline-injected mice. Moreover, direct expression of exogenous STEP61 to enhance its phosphatase activity generated a long-lasting antiallodynic action in CFA-injected mice, suggesting a great contribution of STEP61 dysfunction to the etiology of chronic inflammatory pain.
Taken together, the current study demonstrated an important role of GABAAR/STEP61 signaling in spinal nociceptive processing. Our data shed new light on the potential for STEP61 to treat the diseases relevant to the imbalance between neural excitation and inhibition.
This work was supported by the National Natural Science Foundation of China (grant nos. 31271186 and 31100804), Peking, China.
The authors declare no competing interests.