Activation of NPFFR2 leads to hyperalgesia through the spinal inflammatory mediator CGRP in mice

Experimental Neurology 291 (2017) 62–73

Contents lists available at ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Research Paper

Activation of NPFFR2 leads to hyperalgesia through the spinal inflammatory mediator CGRP in mice Ya-Tin Lin a, Ho-Ling Liu d, Yuan-Ji Day e, Che-Chien Chang h, Po-Hung Hsu f,g, Jin-Chung Chen a,b,c,⁎ a

Graduate Institute of Biomedical Sciences, Department of Physiology and Pharmacology, Chang Gung University, No. 259 Wen-Hwa 1st Rd., Taoyuan, 333, Guishan, Taiwan Healthy Aging Research Center, Chang Gung University, Taoyuan, Taiwan Neuroscience Research Center, Chang Gung Memorial Hospital, No. 5, Fusing St., Taoyuan, 333, Guishan, Taiwan d Department of Imaging Physics, The University of Texas MD Anderson Cancer Center, No. 1400 Pressler St., TX 77030-4009, Houston, USA e Department of Anesthesiology, Hualein Tzu Chi Hospital & Tzu Chi University, Tzu Chi Foundation, No. 707, Chung Yang Rd., 970 Hualien, Taiwan f Department of Electrical Engineering, Chang Gung University, Taoyuan, Taiwan g Center for Advanced Molecular Imaging and Translation, Chang Gung Memorial Hospital, Taoyuan, Taiwan h Department of Chemistry, Fu Jen Catholic University, No. 510, Zhongzheng Road, Xinzhuang Dist. 242, New Taipei City, Taiwan b c

a r t i c l e

i n f o

Article history: Received 11 October 2016 Received in revised form 25 January 2017 Accepted 1 February 2017 Available online 05 February 2017 Keywords: CGRP Hyperalgesia Neuropeptide FF (NPFF) NPFFR2 DRG Pain transmission

a b s t r a c t Neuropeptide FF (NPFF) is recognized as an opioid modulating peptide that regulates morphine-induced analgesia. The aim of this study was to delineate the role of NPFFR2 in pain transmission. We found the expression levels of NPFF and NPFFR2 were increased in the lumbar dorsal horn of animals with CFA- and carrageenan-induced inflammation and both NPFFR2 over-expressing transgenic (NPFFR2-Tg) and NPFFR2 agonist-treated mice displayed hyperalgesia. BOLD signals from functional MRI showed that NPFFR2-Tg mice exhibited increased activation of pain-related brain regions after painful stimulation when compared to WT mice. Inflammatory mediators within the spinal cord, calcitonin gene-related peptide (CGRP) and substance P (SP), were up-regulated in NPFFR2-Tg and chronic NPFFR2 agonist-treated mice. In DRG cultures, treatment with an NPFFR2 agonist induced the expression and release of CGRP, an action which was blocked by NPFFR2 siRNA. Furthermore, treatment with a CGRP antagonist ameliorated the pain hyperalgesia in NPFFR2-Tg mice, returning the pain threshold to a control level. However, treatment with a SP antagonist reduced the pain responses in both WT and NPFFR2Tg mice and did not suppress pain hypersensitivity in NPFFR2-Tg mice. Together, these results demonstrate that NPFFR2 activation modulates pain transmission by up-regulating the pain mediator CGRP, leading to hyperalgesia. © 2017 Elsevier Inc. All rights reserved.

1. Introduction Neuropeptide FF (NPFF; with the sequence FLFQPQRF) belongs to the RF-amide peptide family and was first isolated from bovine brain by virtue of the shared N-terminal sequence, RF-NH2 (Yang et al., 1985; Yang et al., 2008). There are five subgroups of the RF-amide family that include NPFF, prolactin releasing peptide (PrRP), RF-amide related peptide (RFRP), kisspeptin, and pyroglutamylated RF-amide peptide (QRFP) (Jhamandas and Goncharuk, 2013). NPFF is ubiquitously expressed in the central nervous system with highest expression in the hypothalamus, posterior pituitary and spinal cord (Yang et al.,

⁎ Correspondence to: J-C Chen, Graduate Institute of Biomedical Sciences, Department of Physiology and Pharmacology, School of Medicine, Chang Gung University, 259 Wenhua 1st Road, Guishan Dist. 33302, Taoyuan City, Taiwan. E-mail addresses: [email protected] (Y.-T. Lin), [email protected] (H.-L. Liu), [email protected] (Y.-J. Day), [email protected] (C.-C. Chang), [email protected] (P.-H. Hsu), [email protected] (J.-C. Chen).

http://dx.doi.org/10.1016/j.expneurol.2017.02.003 0014-4886/© 2017 Elsevier Inc. All rights reserved.

2008). Two Gi protein-coupled receptors have been identified as NPFF cognate receptors, NPFFR1 (GRP147) and NPFFR2 (GPR74) (Bonini et al., 2000; Elshourbagy et al., 2000). In addition to bind NPFF, NPFFR1 is further recognized as the receptor for RFRP (also known as NPVF) (Hinuma et al., 2000; Liu et al., 2001). NPFFR2 is highly expressed in pain-processing regions such as spinal dorsal horn, thalamus and dorsal raphe nucleus (Liu et al., 2001; Zeng et al., 2003). In DRG neurons, it is synthesized in cell bodies and trans-located to the afferent sensory nerve terminals in the spinal dorsal horn (Gouarderes et al., 2000). The NPFF-NPFFR system is primarily known to function in nociception and opiate analgesia (Panula et al., 1999; Yang et al., 2008). In the spinal cord, these effects are modulated through NPFFR2, since NPFFR1 expression has not been detected (Yang et al., 2008). NPFF-NPFFR signaling results in either pro-nociceptive or antinociceptive effects (Roumy and Zajac, 1998). These opposing outcomes can be specifically stimulated by different routes of NPFF administration (intra-thecal (i.t.) vs. intra-cerebroventricular (i.c.v.)), or regulation via two distinct NPFF receptors (Ayachi and Simonin, 2014; Lameh et al.,

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

2010; Yang et al., 2008). For example, reduced pain threshold in the tailflick test (Yang et al., 1985) and decreased morphine-induced analgesia (Dupouy and Zajac, 1997) was observed in rats after i.c.v. administration of NPFF. I.c.v. injection of an NPFFR2 selective agonist reversed morphine-induced analgesia (Roussin et al., 2005). Neutralization of endogenous NPFF was found to potentiate morphine analgesia (Kavaliers and Innes, 1992; Kavaliers and Yang, 1989). However, i.t. injection with NPFF or its analog produced analgesia or potentiated the opioid analgesic effect (Gouarderes et al., 1996; Kontinen and Kalso, 1995). Nociceptors in the peripheral sensory nerve terminals can be activated by different inflammatory mediators that include cytokines and neurotrophins (Coutaux et al., 2005; Marchand et al., 2005). The ascending nociceptive pathway relies on a cascade of pain mediators, including glutamate, calcitonin gene-related peptide (CGRP) and substance P (SP). CGRP and SP deliver the nociceptive message from DRG to spinal cord neurons and the enhancement of these signals triggers nociceptive ‘central sensitization’ along the pain pathway (Basbaum et al., 2009; Seybold, 2009). These two neurotransmitters are synthesized in the DRG, stored together in the dense-core vehicles of sensory nerve terminals and co-released upon stimulation (Matteoli et al., 1988; Seybold, 2009; Snijdelaar et al., 2000). Two forms of CGRP, αCGRP and βCGRP, which share N 90% sequence homology, bind to calcitonin receptor-like receptor (CLR) (Russell et al., 2014). αCGRP is the predominant form, with expression throughout the nervous system, while βCGRP expression is mostly restricted to the enteric nervous system (Mulderry et al., 1985; Russell et al., 2014). During pain transduction, CGRP is upregulated by neurotrophins, including nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) to modulate the expression of pro-inflammatory cytokines (Lin et al., 2011; Russell et al., 2014). SP is mainly expressed in unmyelinated sensory fibers and like CGRP, its expression is also up-regulated by NGF (Snijdelaar et al., 2000). SP produces a slow excitatory postsynaptic potential and is involved in the transmission of delayed pain signals by binding to neurokinin 1 receptor (Snijdelaar et al., 2000; Zubrzycka and Janecka, 2000). This study focuses on describing the function of NPFFR2 in the spinal cord to delineate its role in nociception. By using genetic and pharmacological tools, we explored a novel mechanism that involved the CGRPleaded pain transmission in NPFFR2-mediated hyperalgesia. 2. Methods 2.1. Animals Male C57BL/6 mice (age 8–10 weeks) were purchased from National Laboratory Animal Center (Taipei, Taiwan) and randomly housed 5–6 per cage in Specific pathogen free (SPF) animal room. Mice were acclimatized to the room with a controlled temperature, air humidity and 12 h day-night cycle (light on at 7:00 AM). Food (Western Lab 7001, Orange, CA, USA) and water were available ad libitum. All the behavioral tests took place during the light cycle. Animal handling and drug treatments were performed in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals, and all procedures were approved by the Animal Care Committee of Chang-Gung University (CGU 08-06 and CGU 13-014). 2.2. NPFFR2 transgenic (Tg) mice The generation of NPFFR2-Tg mice was described previously (Lin et al., 2015). In brief, C57BL/6 NPFFR2 cDNA was cloned into the pWHERE plasmid (InvivoGen, CA, USA) under a 5′-adjacent neuron-specific enolase (NSE) promoter. NPFFR2-Tg mice were generated on a C57BL/6 mice background and maintained as heterozygotes. The formal nomenclature of NPFFR2-Tg mice is C57BL/6C-Tg(Npffr2)Dyj2. Littermate wildtype (WT) mice served as controls. Only male mice age 8–12 weeks were used.

63

2.3. Induction of inflammation in mice Mice were anesthetized with 1.5–2% isoflurane. Complete Freund's adjuvant (CFA, with 0.05 mg heat-killed and dried Mycobacterium tuberculosis, Sigma, St. Louis, MO, USA) was mixed 1:1 with saline to become a water-in-oil emulsion. 30 μl of saline, CFA mixture or 1% carrageenan (Sigma) were injected into both hindpaws of the corresponding groups and lumbar spinal dorsal horn was collected 4, 10, 20, 30 or 72 h after administration for the subsequent PCR quantification. In a different set of experiment, thermal hyperalgesia were measured by the Hargraves' plantar test at time 0 (before administration) and 24, 48, 72 h after the treatment. 2.4. Real-time PCR Total RNA was isolated from tissues using TRIzol® reagent according to the manufacturer's protocol. The mRNA was transcribed into cDNA via reverse transcription PCR (HT BioTechnology, England UK). The cDNA levels for corresponding targets were measured by real-time PCR using the Bio-Rad iCycler PCR Thermal Cycler (Bio-Rad, Hercules, CA, USA) and SYBR (Bio-Rad). The PCR protocol was as follows: 95 °C for 10 min, 95 °C for 15 s and 60 °C for 30 s for 40 cycles. The primer sequences are Rpl35a-forward 5′ GCT GTG GTC CAA GGC CAT TTT 3′; Rpl35a-reverse, 5′ CCG AGT TAC TTT TCC CCA GAT GAC 3′; NPFFR2-forward, 5′ ACA TCT ACC CTT TCG CCC AC 3′; NPFFR2-reverse, 5′ GCT TCT CCC ATT TCC TCT ATC AA 3′; NPFF-forward, 5′ GTA TGC CCA CAT TCC AGA CA 3′; NPFF-reverse, 5′ TGG AGC AGA ACA CGA AAG AG 3′. Threshold cycle, Ct, which correlates inversely with the target mRNA levels, was measured and relative expression levels were calculated. After normalization to Rpl35a, data were expressed as a percentage of the corresponding control. 2.5. Nociceptive tests Mechanical hyperalgesia was measured by the von Frey method. Mice were placed in a clear plexiglass box (L5 × W5 × H10 cm) on an elevated mesh screen. A calibrated von Frey rigid tip (Electronic von Frey Anesthesiometer, IITC Life Science, CA, USA) was applied to the plantar surface of each hindpaw in a series of logarithmically ascending forces. The responses were recorded in grams of paw withdrawal and the average of five applications was referred to as the mechanical hyperalgesia threshold. The interval of each application was 5 min. Thermal hyperalgesia was measured by the hot plate or the Hargraves' plantar test. For the hot plate test, mice were placed on a 50 °C hot plate chamber (Ugo Basile, Varese, Italy) and the first nociceptive response, including lick, raised hindpaw or jump was recorded. For the Hargraves' plantar test, a commercial plantar test instrument was used (Ugo Basile). Mice were placed in the animal enclosure (L10 × W10 × H14 cm) to adapt for 1 h. Infrared intensity was set at 15 and was randomly applied to the mouse plantar surface of each hindpaw. The responses were recorded in seconds of paw withdrawal latency and the average of five applications was referred to as the thermal hyperalgesia threshold. The interval of each application was 5 min. 2.6. Functional MRI (fMRI) Adult male mice (age 10–11 weeks) were initially anesthetized with 1.5% isoflurane. A pair of needle electrodes was inserted subcutaneously into the left forepaw with a distance of 2–3 mm between the two needles for electrical stimulation. A PE10 tube was then placed subcutaneously underneath the back skin for further infusion of dexmeditomidine (Dexdomitor®, Pfizer, New York, NY, USA). Mice then received a subcutaneous injection in the back of 0.05 mg/kg dexmeditomidine. Isoflurane was slowly reduced and stopped within 10 min after the dexmeditomidine injection. The mice were then continuously infused with dexmeditomidine 0.1 mg/kg/h via a syringe

64

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

pump (Harvard PHD 2000, Harvard Apparatus, Holliston, MA, USA) to maintain the anesthesia. During MRI experiments, respiration was monitored by using a pressure-sensitive pad and maintained to 110– 130 breathes per min with 0–0.5% isoflurane in air (Isoflurane was only administered if the breathing rate was over 130 times/min). Body temperature was also monitored by a rectal probe and maintained to 37 ± 0.5 °C using a warm-water circulation. The physiological parameters were recorded continuously by an MR-Compatible Monitoring and Gating System (Model 1025, SA instruments, Stony Brook, NY, USA). At the end of the experiment, mice were intraperitoneally (IP) injected with atipamezole (Antisedan®, Pfizer) at five-times the dosage of dexmeditomidine to reverse the anesthetic drug effect. The duration from the animal preparation to the beginning of the electrical stimulation and fMRI acquisition was kept constant at 40 min to control the anesthesia conditions for all the subjects. The fMRI experiment was performed with electrical stimulation applied to the left forelimb with a rectangular pulse (2 mA, 3 Hz, 0.5 ms) using an Isolated Pulse Stimulator (Model 2100, A-M System, Carlsborg, WA, USA). The left forepaw was stimulated for 60 s, with a 60 s resting period before (baseline) and after the stimulation. The procedure was repeated three times and fMRI data acquisition was continued for another 60 s after the last stimulation block (Fig. 2a–b). The fMRI experiments were conducted on a 7-Tesla Bruker ClinScan 70/30 USR MRI scanner (Ettlingen, Germany) with the BGA 20 gradient. Two actively decoupled radiofrequency coils were used: a volume coil for transmission and a surface array coil positioned on the top of the animal's head as the receiver. A T2*-weighted single-shot gradientecho echo planar imaging (EPI) sequence was used (TR = 2000 ms, TE = 20 ms, FA = 90-degree). A total of ten image slices were imaged in a coronal direction from two separated runs (five slices for each run) with 0.5 mm thickness and in-plane resolution of 0.2 mm × 0.2 mm. Each run was comprised of 300 dynamics with a total acquisition time of 10 mins. The slice coverage included bregma +2.22 to −4.58 (Fig. 2c). Image data from the fMRI scans were processed using SPM2 software (Wellcome Department of Cognitive Neurology, Institute of Neurology, London, UK) and customized codes implemented in the MATLAB version 7.0 (The MathWorks, Natick, MA, USA). Voxels with significant activation were detected by modeling the experimental conditions using a boxcar function convolved with a hemodynamic response function in the context of the general linear model (GLM), with a statistical threshold of 4.41 (p b 0.05, corrected), and overlaid on T2-weighted images for each animal. Activation in five regions of brain, including somatosensory cortex, cingulate cortex, insular cortex, thalamus and PAG, were further evaluated for extent (number of voxels) and the signal change of the detected BOLD response.

2.7. High performance liquid chromatography-electrochemical detection (HPLC-ECD) Animals were sacrificed by decapitation, lumbar spinal dorsal horn was collected and frozen immediately to −80 °C until further analysis. Tissue samples were sonicated in 0.1 N HClO4 on ice and centrifuged at 11,600 x g for 30 min. Supernatants were collected and filtrated for HPLC-ECD detection. The HPLC-ECD system consisted of a pump 125 (Beckman Coulter, Brea, CA, USA), a 5 μm C-18 column (4.6 mm × 150 mm) (Alltech ApolloTM columns, Grace, Deerfield, IL, USA) and an EC detector (BASi, West Lafayette, IN, USA), and the elutes were analyzed by system Gold software (Beckman Coulter). Mobile phase was 9% methanol, 0.105% glacial acetic acid, 29.9 mM citric acid, 50 mM sodium acetate, 52.5 mM NaOH, 1.85 mM 1-octanesulfonic acid, used after filtered and degassed. Amounts of catecholamines, indoleamine and their metabolites were detected by setting oxidation potential at +650 mV.

2.8. Immunohistochemistry (IHC) and immunocytochemistry (ICC) Mice were anesthetized with 45 mg/kg pentobarbital and then perfused with 4% paraformaldehyde. The spinal cord was then removed and cryo-protected with 20% sucrose in potassium phosphate buffer saline (KPBS) overnight at 4 °C. The spinal cord was sliced into 14 μm sections using a freezing microtome (LEICA CM3050S, Bannockburn, IL, USA). After rinsing with KPBS, sections were incubated with primary antibody in 2% normal serum/0.3% Triton X-100/KPBS at 4 °C for 72 h. The antibodies used were anti-CGRP antibody (1:2000, PC205L, Millipore, Bedford, MA, USA) and anti-SP antibody (1:200, GP14110, Neuromics, Minneapolis, MN, USA). Afterward, sections were incubated with 1:300 secondary antibody for 1 h at room temperature. The secondary antibodies used were Alexa Fluor® 488 anti-guinea pig antibody and Alexa Fluor® 594 anti-rabbit antibody (Jackson ImmunoResearch, West Grove, PA, USA). Sections were then incubated with 5 μg/ml 4′,6-diamidino-2-phenylindole (DAPI, Roche, Switzerland) for 5 min before mounting with glycerol. Signals of immunofluorescent stain were observed by fluorescence microscopy. For ICC, DRG cells were fixed with 2% paraformaldehyde for 30 min at room temperature and rinsed with PBS instead of KPBS. 2.9. ELISA Freshly dissected tissues were collected and sonicated in PBS solution (0.01 M, pH = 7.2) on ice to measure the level of CGRP and SP in the DRG or lumbar spinal dorsal horn. Homogenates were centrifuged at 5000 ×g for 5 min at 4° C and the supernatants were obtained and analyzed immediately. The pellets were re-suspended in PBS and protein concentrations were determined by the Coomassie blue method with bovine serum albumin as standards. Alternatively, supernatants from DRG primary cultures were centrifuged at 5000 × g for 5 min at 4° C. The samples were analyzed immediately according to the manufacturer's protocols by CGRP EIA (Cayman, Ann Arbor, Michigan, USA) or SP EIA (Cayman). 2.10. Western blotting Samples were sonicated in 1% sodium dodecyl sulfate (SDS) and heated to 100 °C for 5 min. Equal amounts of proteins were separated by 10% SDS-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore) using 200 mA for 70 min at 4 °C. After blocking with 5% non-fat milk for 1 h, followed by incubation overnight with specific primary antibody, blots were incubated with the secondary antibody conjugated with HRP for 1 h on the second day. The substrate solution (Western Lightning® plus ECL, PerkinElmer, Waltham, MA, USA) was added and reacted for 1 min and enhanced chemiluminescence signal was detected by ChemiDoc MP Imaging System with Image LabSoftware (Bio-Rad, Hercules, CA, USA). Signals were normalized to β-actin. The antibodies used were anti-β-actin (1:5000, Sigma); anti-μ opioid receptor (MOR) (1:1000, Abcam, Cambridge, UK); anti- transient receptor potential vanilloid receptor subtype 1 (TRPV1) (1:1000, Santa Cruz, Santa Cruz, CA, USA). 2.11. Drug treatment The preparation of AC-263093 and CFMHC were described in our previous publication (Lin et al., 2016). WT mice were IP injected with AC-263093 (2.5, 5 and 10 mg/kg) or vehicle 1 h prior to the nociceptive tests. In the chronic drug treatment group, 20 mg/kg AC-263093 was injected once daily for 20 days. The nociceptive test was performed the next day after the last injection and then the mice were sacrificed by decapitation to collect the DRG and lumbar spinal dorsal horn. For experiments with CGRP/SP antagonist, WT or NPFFR2-Tg mice were IP injected with CGRP antagonist BIBN 4096 (Tocris) 10 mg/kg or SP antagonist CP 99994 (Tocris) 30 min prior to the nociceptive test.

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

2.12. Primary DRG cultures The DRG neurons were prepared as previously reported (Lin et al., 2011). Bilateral lumbar L1-L6 DRGs were dissected from 3 week-old SD rats. DRG tissues were digested with 1 mg/ml collagenase (Sigma) at 37 °C for 30 min and then incubated with 0.25% (v/v) trypsin-EDTA (Biological Industries, Israel) at 37 °C for 30 min. Tissues were centrifuged at 300 x g for 5 min and then re-suspended in DMEM/F12 medium (repeated three times, Biological Industries). Next, tissues were manually triturated approximately sixty times using a flame-polished Pasteur pipette. The dissociated cells were suspended in DMEM/F12 medium containing 10% fetal calf serum (Biological Industries), 100 μg/ml penicillin/streptomycin (Gibco, ThermoFisher, Waltham, MA, USA) and 1 mM sodium pyruvate (Gibco) and plated onto poly-Llysine-coated plates. Medium was replaced the following day with the addition of 10 μM Ara-C (Sigma) and 100 ng/ml NGF (Millipore) and changed every two days thereafter. The drug treatment was started on the third day after the cells were plated. 2.13. Measurement of CGRP and SP release On the third day after cell plating, cultures were changed to 0.2 ml of complete medium for 30 min. Then primary DRG cells were treated with 1 μl dNPA/AC-263093 (0.0005, 0.005, 0.05, 0.5 and 5 nmol) or vehicle for 1 h in a CO2 incubator. dNPA (D.Asn-Pro-(N-Me)Ala-Phe-LeuPhe-Gln-Pro-Gln-Arg- Phe-NH2) was commercially synthesized by Genemed Synthesis (San Antonio, TX, USA) and dissolved in 20% methanol in PBS, which was diluted to 10% methanol before use. AC-263093 was dissolved in 5% DMSO/10% Tween 20 in PBS and diluted to 2.5% DMSO/5% Tween 20 before use. After incubation, the supernatants were collected and centrifuged at 5000 ×g for 5 min to remove the tissue residues before submitting CGRP/SP to ELISA. Pre-treatment with naloxone (5 nmol, Tocris, Bristol, UK) or PBS 15 min prior to the addition of dNPA (5 nmol) for 1 h was used to test the involvement of MOR. For the RNA interference test, siRNA was transfected on the third day of cell plating and incubated for another 72 h prior to dNPA (5 nmol) treatment for 1 h. NPFFR2 siRNA (On-TARGETplus SMARTpool siRNA, Dharmacon, Lafayette, CO, USA) or the non-targeting control pool siRNA 20 nM was transfected using the NeuroPORTER™ transfection reagent (Genlantis, San Diego, CA, USA) according to the manufacturer's protocols. 2.14. Data analyses and statistical analysis ImageJ (NIH, Bethesda, MD, USA) software was used to quantify the immuno-staining images. A threshold was set by Image J to define the positive staining among different experiments of the same study. For spinal cord immuno-staining, six different sections (spacing 80– 90 μm) were selected and the positive area above the set threshold was calculated. All data were analyzed with the GraphPad InStat (GraphPad Software, San Diego, CA, USA) or the GraphPad Prism 5 (GraphPad). Results are expressed as mean ± S.E.M. The data were analyzed by one-way ANOVA followed by the Tukey multiple comparisons test, two-way ANOVA followed by the Bonferroni multiple comparisons test or the un-paired Student's t-test. A value of p b 0.05 was considered statistically significant. 3. Results 3.1. Up-regulation of spinal NPFF and NPFFR2 in mouse models of inflammatory pain In order to validate the involvement of the NPFF-NPFFR2 system in inflammation-induced pain, mRNA levels of spinal cord NPFF and NPFFR2 were measured in the complete Freund's adjuvant (CFA)- and carrageenan-induced inflammation mouse models. Lumbar spinal

65

dorsal horn was collected at time points ranging from 4 to 72 h after bilateral CFA or carrageenan intra-hindpaw injection. Examples of swollen hindpaws after treatment are shown in Fig. S1. mRNA expression of NPFF increased in mice 10–72 h following the injection with CFA and peaked at 20 h after the injection with carrageenan (Fig. 1a). Two-way ANOVA indicated statistical significance between treatment groups (F(2, 4) = 9.147, p = 0.0003) and treatment × duration interaction (F(2, 4) = 4.498, p = 0.002). A time-dependent profile of NPFFR2 mRNA expression was also measured after intra-hindpaw CFA and carrageenan injection (Fig. 1b). Two-way ANOVA indicated a statistical significance between treatment groups (F(2, 4) = 6.457, p = 0.0029) and the treatment × duration interaction (F(2, 4) = 2.246, p = 0.0034). The results indicate that spinal cord NPFF and NPFFR2 are up-regulated in models of inflammatory pain. 3.2. NPFFR2-Tg mice exhibited mechanical and thermal hyperalgesia NPFFR2-Tg mice were subjected to thermal and mechanical pain tests to clarify the function of NPFFR2 on nociception. The thermal and mechanical pain thresholds were significantly reduced in the NPFFR2Tg mice when compared to WT mice (Fig. 1c and d) (Hot plate, p = 0.0347, t(10) = 2.442, unpaired t-test; von Frey, p = 0.0102, t(24) = 2.789, unpaired t-test). To examine the role of NPFFR2 on chronic inflammatory pain, CFA and carrageenan were administered in mice hindpaw and pain responses were recorded at 0, 24, 48 and 72 h after injection. The nociceptive responses were increased in a time-dependent manner in both WT and NPFFR2-Tg mice after bilateral intrahindpaw CFA injection as compared to saline-treated group (Fig. 1e). Alternatively, the nociceptive responses reached a maximal level at 24 h after bilateral intra-hindpaw carrageenan injection and then subsided gradually (Fig. 1f). Nevertheless, in both CFA and carrageenan treatment, NPFFR2-Tg mice displayed enhanced pain responses comparing to WT mice. Two-way ANOVA indicated statistical significance between WT and NPFFR2-Tg groups after CFA treatment (Saline, F(1, 3) = 31.31, p b 0.0001; CFA, F(1, 3) = 24.03, p b 0.0001) or after carrageenan treatment (Saline, F(1, 3) = 23.28, p b 0.0001; Carrageenan, F(1, 3) = 36.97, p b 0.0001). 3.3. Hyperreactivity of the pain transmission pathway in NPFFR2-Tg mice Painful stimuli-induced brain activation was evaluated by fMRI. The neural activity of brain regions that are associated with pain transmission such as the somatosensory cortex, cingulate cortex, insular cortex, thalamus and periaqueductal gray (PAG) was measured following forepaw electrical stimulation. A schematic diagram of electrical stimulation, BOLD response and regions of interest are shown in Fig. 2a–c. NPFFR2-Tg mice exhibited enhanced neural activation after electrical stimulation in these pain-related brain regions when compared to WT mice (Fig. 2d). Interestingly, BOLD signal changes in the PAG were observed in NPFFR2-Tg, but not WT mice. The change of BOLD signal intensity averaged over three repetitive stimulations in the left and right brain are shown in Fig. 3a-h. The extent (number of voxels) of detected activation was increased in NPFFR2-Tg mice when compared with WT mice and reached a significant level in the right cingulate cortex (Fig. 3i) (p = 0.0179, t(10) = 2.771, unpaired t-test). The extent of activation in the right sensory and insular cortex, left cingulate cortex and PAG of NPFFR2-Tg mice increased in NPFFR2-Tg mice, however did not reached statistical significance when compared with WT controls (p values of 0.09, 0.07, 0.06 and 0.07, respectively). Measurements of the average signal intensity (SI) produced similar results when comparing NPFFR2-Tg to WT mice (Fig. 3j). The BOLD signal intensity significantly increased in the right insular cortex (p = 0.0126, t(10) = 3.032, unpaired t-test), left cingulate cortex (p = 0.0493, t(10) = 2.2362, unpaired t-test) and thalamus (p = 0.0405, t(10) = 2.2353, unpaired t-test) of NPFFR2-Tg mice. The BOLD signal intensity increased in the left sensory cortex and PAG, but did not reached the

66

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

Fig. 1. Up-regulation of spinal cord NPFF and NPFFR2 after carrageenan or CFA treatment in WT mice and pain responses of WT and NPFFR2-Tg mice. (a, b) Gene expression profiles of NPFF and NPFFR2 in the lumber spinal dorsal horn of mice with carrageenan or CFA-induced inflammation for 4, 10, 20, 30, 72 h. The results were normalized to the saline control. Data are expressed as mean ± S.E.M. and analyzed using two-way ANOVA with Bonferroni post-hoc tests. *p b 0.05, ***p b 0.001; compared to saline control (N = 5 per group). (c-f) The Hot plate, Von-Frey and Hargreaves' plantar test were used to evaluate the thermal and mechanical nociceptive responses in WT and NPFFR2-Tg mice. (c, d) Nociceptive response to a thermal (N = 4 for WT mice, N = 8 for Tg mice) and mechanical stimulus (N = 10 for WT mice, N = 16 for Tg mice). Data are expressed as mean ± S.E.M. and analyzed by an un-paired Student's t-test. *p b 0.05; compared to WT controls. (e, f) The thermal threshold was recorded at different time points (0, 24, 48, 72 h) after bilateral intra-hindpaw saline, CFA or carrageenan injection (N = 5–6 per group). Data are expressed as mean ± S.E.M. and analyzed by two-way ANOVA with Bonferroni post-hoc tests. ***p b 0.001; compared to WT controls.

statistical significance when compared to the WT mice (p value of 0.07 and 0.064, respectively). Overall, the fMRI results indicate that after nociceptive stimulation, NPFFR2-Tg mice exhibit stronger brain activity along the pain transmission pathway when compared to WT mice. Because the activation of PAG in NPFFR2-Tg mice was observed, we further evaluated activation of the pain descending pathway by measuring monoamines in the dorsal spinal cord with HPLC-ECD. The levels of norepinephrine (NE), serotonin (5HT) and the NE metabolite, 3-methoxy-4-hydroxyphenylglycol (MHPG), showed no difference between WT and NPFFR2-Tg mice (Fig. S2a). However, the 5HT metabolite, 5-hydroxyindoleacetic acid (5HIAA), was significantly increased in the spinal dorsal horn of NPFFR2-Tg mice (Fig. S2a) (p = 0.0247, t(7) = 2.849, unpaired t-test). In this context, there was no observable change in NE turnover rate when comparing WT and NPFFR2-Tg mice, the turnover rate of 5HT was increased in the dorsal horn of NPFFR2-Tg mice (Fig. S2b) (p = 0.0041, t(7) = 4.190, unpaired t-test). 3.4. Increased level of CGRP and SP in the lumbar spinal dorsal horn and DRG of NPFFR2-Tg mice Pain mediators were measured in the spinal cord and DRG of NPFFR2-Tg and WT mice. The protein expression of CGRP and SP were up-regulated in the lumbar spinal dorsal horn of NPFFR2-Tg mice when compared to WT mice as measured by immunohistochemistry (Fig. 4a–f) (CGRP, p = 0.0388, t(8) = 2.468; SP, p = 0.0626, t(8) = 2.162; SP, unpaired t-test). CGRP protein significantly increased in the DRG and the lumbar spinal dorsal horn of NPFFR2-Tg mice when compared to WT mice (DRG, p = 0.0239, t(6) = 3.002; lumbar spinal dorsal horn, p = 0.0483, t(6) = 2.472, unpaired t-test), while the amount of SP did not show statistical significance (Fig. 4g and h) as quantified

by ELISA. Primary DRG neurons from WT and NPFFR2-Tg mice were stained with anti-CGRP or -SP antibodies. CGRP and SP were expressed in the DRG neurons and were increased in the cell body and nerve fibers of NPFFR2-Tg mice when compared to WT mice (Fig. S3). The μ opioid receptor (MOR) decreased in the spinal cord of NPFFR2-Tg mice but the amount of transient receptor potential vanilloid receptor subtype 1 (TRPV1) was not changed in the two groups of mice (Fig. 5) (MOR, p = 0.0016, t(22) = 3.598, unpaired t-test). 3.5. Induction of hyperalgesia and up-regulation of CGRP and SP by treatment with non-peptide NPFFR2 agonist in WT mice Two non-peptide NPFFR2 agonists were utilized (AC-263093 and CFMHC) (Gaubert et al., 2009; Lameh et al., 2010) to confirm the pro-nociceptive role of NPFFR2 on pain transmission observed in the NPFFR2-Tg mice. WT mice were treated with AC-263093 (2.5– 10 mg/kg) for evaluating the acute nociceptive responses. The thermal nociceptive threshold was measured by the Hargreaves' test after drug treatment for 1 h. The activation of the NPFFR2 induced a dosedependent pain response (Fig. 6a). The data were analyzed using oneway ANOVA and exhibited statistical significance (F(3, 16) = 7.872, p = 0.0019). Tukey's multiple comparison tests showed significant increase in the dosage of 5 mg/kg (p b 0.05) and 10 mg/kg (p b 0.01). When AC-263093 was applied for 20 days to WT mice, they exhibited thermal hyperalgesia when compared to vehicle controls (Fig. 6b) (p = 0.0310, t(9) = 2.553, unpaired t-test). CGRP protein was increased in the DRG and lumbar spinal dorsal horn after administration of AC263093 for 20 days (Fig. 6c) (DRG, p = 0.0007, t(10) = 4.836; lumbar spinal dorsal horn, p = 0.0456, t(8) = 2.365, unpaired t-test). AC263093 treatment increased the level of SP in the DRG, but not the lumbar spinal dorsal horn (Fig. 6d) (DRG, p = 0.0461, t(10) = 2.276,

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

67

Fig. 2. Electrical stimulation paradigm and brain fMRI response. 7 T-fMRI was used to detect the brain activity evoked by electrical stimulation. (a) Schedule of three repetitive electrical stimulations. (b) Time course of BOLD signals following electrical stimulation obtained from cingulate cortex of a NPFFR2-Tg mouse. Three cycles of resting and stimulation blocks were separated by a dotted line with gray areas indicating the ON-period of the stimulation. (c) The mouse brain map illustrates the position of ten scanned sections with 0.5 mm thickness separated by a 0.2 mm gap. (d) Color overlay of brain activation on T2-weighted images of designated brain regions in WT and NPFFR2-Tg mice. Intensity is increased from yellow to red. Sensory: sensory cortex; Cingulate: cingulate cortex; Insular: insular cortex. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

unpaired t-test). Of note, levels of NPFFR2 remained unchanged in the lumbar spinal dorsal horn of WT mice after chronic AC-263093 treatment (Fig. S4). The role of NPFFR2 on inflammatory pain was confirmed by CFMHC. WT mice exhibited thermal hyperalgesia (Fig. S5a) and increased the expression of CGRP in the DRG, but not the dorsal spinal cord and had no effect on the level of SP following injection of CFMHC for 20 days (Fig. S5b–c). 3.6. NPFFR2 agonist-induced release and expression of CGRP/SP in DRG primary cultures DRG primary cultures were used to examine if NPFFR2 activation could directly cause CGRP and/or SP release. CGRP and SP were released from the DRG cultures in a plausible dose-dependent manner after treatment for 60 min with the dNPA agonist (Fig. 7a and b). The data

were analyzed by one-way ANOVA and showed statistical significance for either CGRP (F(5, 34) = 5.134, p = 0.0013) or SP (F(5, 37) = 10.971, p b 0.0001). An increase of CGRP and SP originating from the cultured DRG neurons was also observed following treatment with 5 nmol AC-263093 (Fig. 7c and d) (CGRP, F(5, 24) = 2.798, p = 0.0396; SP, F(5, 28) = 6.389, p = 0.0004, one-way ANOVA). Treatment of AC-263093 (5 nmol for 24 or 48 h) increased the expression of CGRP, but not SP in DRG cultures (Fig. 7e) (CGRP, F(2, 23) = 8.325, p = 0.0019, one-way ANOVA). The MOR antagonist, naloxone, was used to examine the involvement of opiate receptor on NPFFR2-induced CGRP/SP release. dNPAinduced CGRP and SP release was unaffected by pre-treatment of naloxone (5 nmol) for 15 min (Fig. 8a and b). This result indicates that NPFFR2-mediated CGRP and SP release does not require the activation of MOR.

68

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

Fig. 3. Brain activity of WT and NPFFR2-Tg mice following electrical stimulation. Functional MRI was used to validate the difference of brain activity under painful stimulation between WT and NPFFR2-Tg mice. (a-h) Relative BOLD signals from an average of three repetitive electrical stimulations recorded in the designated brain regions. Quantitative evaluation includes the activation extent (i) and BOLD signal intensity (SI) change (j) (N = 6 per group). Data are expressed as mean ± S.E.M. and analyzed by an un-paired Student's t-test. *p b 0.05; compared to WT controls. Sensory: sensory cortex; Cingulate: cingulate cortex; Insular: insular cortex. L: lift brain; R: right brain. SI: signal intensity.

Fig. 4. Immunoreactivity of CGRP and SP in the lumbar spinal dorsal horn and DRG of WT and NPFFR2-Tg mice. Immunofluorescence was used to measure the CGRP (a–b) and SP (c–d) expression in the lumbar spinal dorsal horn of WT and NPFFR2-Tg mice. (e, f) CGRP and SP expression was evaluated by quantification of the corresponding immuno-reactive (ir) area of the dorsal spinal cord (N = 5 per group). (g, h) CGRP and SP expression was measured in the DRG and lumbar spinal dorsal horn of WT and NPFFR2-Tg mice by ELISA (N = 4 per group). Data are expressed as mean ± S.E.M. and analyzed by an un-paired Student's t-test. *p b 0.05; compared to WT controls. Scale bar, 100 μm. SC: lumbar spinal dorsal horn.

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

69

The NPFFR2 siRNA was added to DRG cultures to validate the NPFFR2 effect on the release of CGRP and SP. NPFFR2 siRNA was transfected in DRG cultures 72 h prior to dNPA (5 nmol) administration for 1 h to stimulate the release of CGRP and SP. Pre-treatment of NPFFR2 siRNA decreased the effect of dNPA on the release of CGRP, but not SP (Fig. 8c and d). 3.7. Effect of CGRP and SP antagonists on the nociceptive responses of WT and NPFFR2-Tg mice CGRP and SP antagonists were utilized to confirm the involvement of these pain mediators in NPFFR2-mediated pain hyperalgesia. Treatment of CGRP antagonist (BIBN 4096, 10 mg/kg) for 30 min significantly restored the pain responses of NPFFR2-Tg mice (Fig. 9a) (p b 0.01, t = 3.730, two-way ANOVA, Bonferroni multiple comparisons test). The pain response level of BIBN 4096-treated NPFFR2-Tg mice was similar to BIBN 4096-treated WT mice. Treatment of SP antagonist (CP 99994, 10 mg/kg) for 30 min significantly reduced the pain responses in both WT and NPFFR2-Tg mice (Fig. 9b) (WT mice, p b 0.001, t = 5.072; NPFFR2-Tg mice, p b 0.001, t = 4.468, two-way ANOVA, Bonferroni multiple comparisons test). However, NPFFR2-Tg mice still showed pain hypersensitivity in either vehicle or CP 99994-treated groups when compared to corresponding WT mice (vehicle, p b 0.05, t = 3.054; CP 99994, p b 0.01, t = 3.224, two-way ANOVA, Bonferroni multiple comparisons test). Fig. 5. The level of pain modulatory proteins in the lumbar spinal dorsal horn of WT and NPFFR2-Tg mice. MOR and TRPV1 were resolved and quantified by western immunoblot. Gel bands illustrate the representative pictures presented in duplicate. Quantitative data were normalized to β-actin. Data are expressed as mean ± S.E.M. and analyzed by an un-paired Student's t-test. **p b 0.01; compared to corresponding WT controls (N = 11–13 per group).

4. Discussion Herein, we report a novel modulatory role of NPFFR2 on pain transmission and hyperalgesia. NPFF and NPFFR2 mRNA were up-regulated in the spinal cord of mice with inflammatory pain. Activation of NPFFR2 facilitated pain transmission and potentiated the algesic response, which was evidenced by both in vivo and in vitro studies.

Fig. 6. Nociceptive response and protein levels of CGRP and SP after NPFFR2 agonist (AC-263093) treatment in WT mice. (a) Thermal nociceptive responses were evaluated by Hargreaves' plantar test after treatment of various doses of NPFFR2 agonist (AC-263093, IP) for 1 h. Drug dosage zero indicates vehicle group. Data are expressed as mean ± S.E.M. and analyzed by oneway ANOVA with Tukey's multiple comparison tests. *p b 0.05, **p b 0.01; compared to vehicle controls (N = 5 per group). (b) Thermal nociceptive responses were evaluated by Hargreaves' plantar test after daily treatment for 20 days with the NPFFR2 agonist (AC-263093, 20 mg/kg, IP) (N = 5 for vehicle, N = 6 for AC-263093) (c, d) The levels of CGRP and SP in the DRG and lumbar spinal dorsal horn were measured by ELISA (DRG, N = 6 per group; lumbar spinal dorsal horn, N = 5). Data are expressed as mean ± S.E.M. and analyzed by an un-paired Student's t-test. *p b 0.05, ***p b 0.001; compared to vehicle controls. SC: lumbar spinal dorsal horn.

70

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

Fig. 7. NPFFR2 agonist-induced release and expression of CGRP and SP in DRG primary cultures. (a-d) The release of CGRP and SP from the DRG primary cells was evaluated after treatment of various doses of the selective NPFFR2 agonist dNPA or non-peptide agonist AC-263093 for 1 h. Data are expressed as mean ± S.E.M. and analyzed by one-way ANOVA with Tukey's multiple comparison tests. *p b 0.05, **p b 0.01, ***p b 0.001; compared to vehicle controls (N = 4–12 per group). (e) CGRP and SP levels after treatment with AC-263093 (1 μM) for 24 or 48 h in the DRG cells were analyzed by ELISA. Data are expressed as mean ± S.E.M. and analyzed by the one-way ANOVA with Tukey's multiple comparison tests. **p b 0.01; compared to vehicle controls (N = 8 for time point 0, N = 9 for time point 24 and 48 h).

NPFFR2-Tg mice exhibited increased neuronal activity along pain ascending pathway after hindpaw electrical stimulation as observed by fMRI. The pain mediators CGRP and SP, were up-regulated in the DRG and the spinal dorsal horn of NPFFR2-Tg mice and chronic NPFFR2 agonists-treated WT mice. In DRG cultures, NPFFR2 activation induced the expression and release of CGRP. Systemic treatment with a CGRP

antagonist ameliorated pain hyperalgesia in NPFFR2-Tg mice. The release of CGRP from DRG to spinal dorsal horn appears to be responsible for NPFFR2-mediated hyperalgesia. CFA and carrageenan were used to confirm that expression of NPFFNPFFR2 is up-regulated in mice spinal cord during inflammatory pain. The different time profiles of the increase in NPFF vs. NPFFR2 mRNA

Fig. 8. Treatment of MOR antagonist or NPFFR2 siRNA on the release of dNPA-induced CGRP and SP in DRG primary cultures. (a, b) The release of CGRP and SP from DRG cultures after pretreatment with naloxone (5 nmol) 15 min prior to the addition of dNPA (5 nmol) for 1 h. Data are expressed as mean ± S.E.M. and analyzed by one-way ANOVA with Tukey's multiple comparison tests. *p b 0.05, **p b 0.01; compared to vehicle controls. ##p b 0.01, ###p b 0.001; compared the treatment of naloxone + dNPA with naloxone alone (N = 5 per group). NAX: naloxone. (c, d) NPFFR2 siRNA or control siRNA was transfected into DRG cultures and incubated for 72 h. dNPA (5 nmol) was then applied for 1 h to induce the CGRP and SP release. Data are expressed as mean ± S.E.M. and analyzed by two-way ANOVA with Bonferroni post-hoc tests. **p b 0.01, ***p b 0.001; compared to corresponding vehicle controls (N = 12 per group).

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

71

Fig. 9. Effect of CGRP and SP antagonist on pain responses of WT and NPFFR2-Tg mice. Thermal nociceptive responses were evaluated by Hargreaves' plantar test after systemic treatment of CGRP or SP antagonists in WT and NPFFR2-Tg mice. (a) Pain responses after mice were treated with CGRP antagonist (BIBN 4096, 10 mg/kg, IP) or vehicle for 30 min. (b) Pain responses after mice were treated with SP (CP 99994, 10 mg/kg, IP) or vehicle for 30 min. Data are expressed as mean ± S.E.M. and analyzed by two-way ANOVA with Bonferroni post-hoc tests. *p b 0.05, **p b 0.01; compared to WT mice. ##p b 0.01, ###p b 0.001; compared to vehicle controls (N = 7 for WT controls of treatment with CGRP antagonist, N = 6 for WT controls of treatment with SP antagonist, N = 5 for NPFFR2-Tg mice). n.s. = no significant difference.

might result from the distinct chemical nature of CFA and carrageenan and induced inflammation process. Carrageenan is a mucopolysaccharide derived from Chondrus and was known to produce acute inflammatory response with significant tissue edema (Winter et al., 1962). CFA contains heat-killed Mycobacterium tuberculosis and was known to produce choric inflammatory pain with longer on-set and duration than carrageenan (Honor et al., 1999; Ren and Dubner, 1999). The upregulation of NPFF mRNA was consistent with the processes of carrageenan- or CFA-induced inflammatory symptoms. On the other hand, a seemly negative correlation was observed between NPFF and NPFFR2 gene regulation. Though difficult to speculate, it might account as a negative feedback, or independent regulation, between ligand and receptor. The increase of NPFF-NPFFR2 is consistent with previous studies that show up-regulation of NPFF and NPFFR2 in the spinal cord occurs in animal models of colonic inflammation or carrageenaninduced hyperalgesia (Nystedt et al., 2004; Vilim et al., 1999; Yang and Iadarola, 2003) and the increased binding of [125I]-1DMe-NPFF in spinal cord during Freund's adjuvant induced-joint inflammation (Lombard et al., 1999). The involvement of the NPFF-NPFFR system in pain regulation has been explored for more than thirty years, but it has been difficult to distinguish between the actions of NPFFR1 and NPFFR2. Thus, NPFFR2 over-expressing mice have provided a unique tool for pinpointing the role of NPFFR2 in nociception. In the present study, NPFFR2-Tg mice showed pain hypersensitivity. Consequent up-regulation of CGRP in the sensory neurons of NPFFR2-Tg mice further supports the hypothesis that NPFFR2 is involved in pro-nociceptive transmission. Chronic treatment with AC-263093 not only reduced the pain threshold, and elevated CGRP expression in the DRG and spinal cord. These findings were supported by the use of another NPFFR2 agonist, CFMHC, which showed similar effects on pain response and CGRP expression. CGRP and SP are robustly expressed in primary sensory neurons and approximately 70% of each protein are co-stored in vesicles at the nerve terminals within the spinal cord (Seybold, 2009). Inflammatory signals, neurotrophins or noxious stimuli are known to increase the expression of CGRP and SP in the DRG neurons (Basbaum et al., 2009; Lin et al., 2011; Seybold, 2009). The NPFFR2 high binding affinity and selective agonist dNPA (Roussin et al., 2005), increased the release of CGRP and SP in a dose-dependent manner in DRG cultures. This result suggests that the modulatory role of the NPFF-NPFFR2 system on pain transmission is distinct from a previously well-described anti-opioid mechanism (Gouarderes and Zajac, 2007; Roumy et al., 2007). Furthermore, the effect of NPFFR2 activation on CGRP appears to be more robust than the effect on SP both in vivo and in vitro. For example, in animal models, regulation of SP by NPFFR2 activation was not as apparent as CGRP and in vitro the knock-down of NPFFR2 by siRNA showed no effect on the release of SP, suggesting that dNPA induction

of SP does not occur through NPFFR2. CGRP and SP antagonists were used to validate their involvement in NPFFR2-mediated hyperalgesia. CGRP antagonist dose-dependently decreased the CFA-induced pain responses (Hirsch et al., 2013). SP antagonist was also reported to attenuate mechanical and thermal hypersensitivity in a rat pain model (Vera-Portocarrero and Westlund, 2004). In the current study, both CGRP and SP antagonists significantly ameliorated the hyperalgesia in NPFFR2-Tg mice. CGRP antagonist totally recovered pain response of NPFFR-Tg mice to a level similar to WT mice. Based on in vitro and in vivo findings, the results indicate NPFFR2-induced hyperalgesia mainly relies on CGRP-mediated effects. This is evidenced by continued pain hypersensitivity in NPFFR2-Tg mice after treatment with SP antagonist. In light of our in vitro study, the data suggest that dNPA-stimulated SP release from DRG was an indirect effect mediated by NPFFR2. Current functional and anatomical brain imaging methods can provide real-time observations of neural activity in live animals which allowed us to probe specific regions known to constitute the neural pathways of pain transmission (Borsook and Becerra, 2011). The major obstacles to using fMRI are sensitivity and resolution. Anesthesia reduces BOLD signals, and animal movement upon electrical stimulation triggers physiological noise that interferes with the recruitment of images (Williams et al., 2010). Medetomidine, a α-agonist, was chosen for fMRI testing after comparison with isoflurane and α-chloralose. The electrical forepaw stimulation has been widely used to assess the BOLD signals associated with pain (Adamczak et al., 2010; Bosshard et al., 2010). The stimulation amplitudes of 1.5 mA and 2 mA are considered as noxious and painful stimulation and 0.5 mA is considered to be an innocuous stimulation (Bosshard et al., 2010). Using these parameters, we detected neuronal activity in response to nociceptive simulation and results showed that left forepaw stimulation produced BOLD signals in bilateral brain regions of the pain pathway, including the somatosensory cortex, thalamus, insular and cingulate cortex. Insular and cingulate cortex function to integrate pain signals with cognitive substrates after stimulation (Porro et al., 2002; Shackman et al., 2011). The increase of BOLD signals indicating enhanced pain transmission in NPFFR2-Tg mice suggests that NPFFR2 up-regulation triggered central sensitization to painful stimulation. Additionally, an increase of BOLD signal in the PAG and an enhanced serotonin turnover rate in dorsal spinal cord, suggesting the activation of the pain descending pathway in NPFFR2-Tg mice. We speculate that activation of the pain-inhibiting pathway may occur as a compensatory effect of augmented pain sensation in NPFFR2-Tg mice. The brain NPFF-NPFFR system is functionally linked with the opioid system (Mouledous et al., 2010), although NPFF displays no binding affinity towards opioid receptors (Gouarderes et al., 1998). Functional interplay between NPFF and the opioid system has been postulated and include MOR internalization (Roumy et al., 2007), MOR/NPFFR2

72

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73

heterodimer formation (Carayon et al., 2014), GRK2-mediated transphosphorylation (Mouledous et al., 2012) or involvement of voltagegated Ca2 + channels (Rebeyrolles et al., 1996; Roumy and Zajac, 1999). There is a general consensus that i.c.v. administered NPFF induces pro-nociception effects, while i.t. administered NPFF produces anti-nociception effects (Yang and Iadarola, 2006; Yang et al., 2008). Nevertheless, an anti-opioid action of NPFFR2 in rat spinal cord is apparent from the observation that the selective NPFFR2 agonist dNPA reduces DAMGO (μ opioid agonist)-stimulated [35S]GTPγS binding (Gouarderes and Zajac, 2007). In the present study, we identified a novel role of NPFFR2 in peripheral neurons that is associated with pain transmission. The effects of NPFFR2 agonists-induced CGRP release from DRG cells seems to be independent of MOR activity, as evidenced by the use of naloxone. However, the level of MOR in spinal dorsal horn of NPFFR2-Tg mice was decreased comparing to WT mice. Hence, the influence of MOR signals in NPFFR2-mediated CGRP release might not be completely excluded. Besides NPFF, other RF-amide peptides that exhibit binding affinity towards NPFFR2, including RFRP-3, PrRP20, kisspeptins-10 and 26RFa are all known to increase pain response (Elhabazi et al., 2013; Yang et al., 1985). The effect of these peptides in NPFFR2-mediated hyperalgesia is unknown and requires further investigation. AC-263093 activates NPFFR2 and is known to modulate pain sensitivity (Gaubert et al., 2009). AC-263093 was chosen because of its selectivity for NPFFR2. The pEC50 value determined by a R-SAT assay is 5.9 ± 0 for NPFFR2 (% efficacy of 90 ± 15 to NPFF), while for NPFFR1 pEC50 value was unable to be determined (% efficacy of 12 ± 0 to NPAF) (Lameh et al., 2010). Similar results were achieved with a cAMP functional assay. An additional compound (CFMHC) was tested which exhibits 114% efficacy towards NPFFR2 and 46% towards NPFFR1 by the R-SAT assay (Gaubert et al., 2009). Both drugs produced similar effects with regard to modulation of pain response and the induction of CGRP expression. AC-263093 is also an NPFFR1 antagonist (Malin et al., 2015). The effects of these drugs are speculated to be mediated through NPFFR2 because there is no detectable expression of NPFFR1 in the spinal cord (Gouarderes et al., 2004b; Gouarderes et al., 2002). The fact that dNPA and AC-263093 increased the CGRP and SP release in DRG cultures supports the in vivo findings that the effect of AC-263093 is mediated by NPFFR2. Inconsistent results have been reported, i.e. AC-263093 was shown to produce anti-nociceptive effects in rats (Lameh et al., 2010). The dosage of AC-263093 in the experiments was comparable to that in the current experiments (1–10 mg/kg vs. 2.5–10 mg/kg). Thus the discrepancy may be best explained by the difference in species employed in each study. AC-263093 is known to be a mixed NPFFR2 agonist and NPFFR1 antagonist, and the relative expression of NPFFR1 and NPFFR2 in different rodent species is quite variable (Gouarderes et al., 2004a; Malin et al., 2015). Importantly, our data were validated with multiple NPFFR2 agonists to ensure repeatability of results. Using genetic and pharmacological tools in combination with functional brain imaging and DRG cultures, we uncovered a novel role and mechanism for NPFFR2 in pain transmission. The findings demonstrated that NPFFR2 plays a significant role in hyperalgesia and possibly exerts its effects on pain modulation through CGRP release from the DRG. Conflict of interest The authors declare no conflict of interests. Acknowledgments We thank Dr. M. Calkins and Dr. A. Stern for English editing and Center for Advanced Molecular Imaging and Translation, Chang Gung Memorial Hospital for equipment support. This work was supported by the Chang Gung Memorial Hospital (CMRPD1C0523, CRRPD1C0023, CMRPD1D0293), Chang Gung University, Healthy Aging Research

Center (EMRPD1E1641) and Ministry of Science and Technology (1042320-B-182-009). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.expneurol.2017.02.003.

References Adamczak, J.M., Farr, T.D., Seehafer, J.U., Kalthoff, D., Hoehn, M., 2010. High field BOLD response to forepaw stimulation in the mouse. NeuroImage 51, 704–712. Ayachi, S., Simonin, F., 2014. Involvement of mammalian RF-amide peptides and their receptors in the modulation of nociception in rodents. Front. Endocrinol. 5, 158. Basbaum, A.I., Bautista, D.M., Scherrer, G., Julius, D., 2009. Cellular and molecular mechanisms of pain. Cell 139, 267–284. Bonini, J.A., Jones, K.A., Adham, N., Forray, C., Artymyshyn, R., Durkin, M.M., Smith, K.E., Tamm, J.A., Boteju, L.W., Lakhlani, P.P., Raddatz, R., Yao, W.J., Ogozalek, K.L., Boyle, N., Kouranova, E.V., Quan, Y., Vaysse, P.J., Wetzel, J.M., Branchek, T.A., Gerald, C., Borowsky, B., 2000. Identification and characterization of two G protein-coupled receptors for neuropeptide FF. J. Biol. Chem. 275, 39324–39331. Borsook, D., Becerra, L., 2011. CNS animal fMRI in pain and analgesia. Neurosci. Biobehav. Rev. 35, 1125–1143. Bosshard, S.C., Baltes, C., Wyss, M.T., Mueggler, T., Weber, B., Rudin, M., 2010. Assessment of brain responses to innocuous and noxious electrical forepaw stimulation in mice using BOLD fMRI. Pain 151, 655–663. Carayon, K., Mouledous, L., Combedazou, A., Mazeres, S., Haanappel, E., Salome, L., Mollereau, C., 2014. Heterologous regulation of Mu-opioid (MOP) receptor mobility in the membrane of SH-SY5Y cells. J. Biol. Chem. 289, 28697–28706. Coutaux, A., Adam, F., Willer, J.C., Le Bars, D., 2005. Hyperalgesia and allodynia: peripheral mechanisms. Joint Bone Spine 72, 359–371. Dupouy, V., Zajac, J.M., 1997. Neuropeptide FF receptors control morphine-induced analgesia in the parafascicular nucleus and the dorsal raphe nucleus. Eur. J. Pharmacol. 330, 129–137. Elhabazi, K., Humbert, J.P., Bertin, I., Schmitt, M., Bihel, F., Bourguignon, J.J., Bucher, B., Becker, J.A., Sorg, T., Meziane, H., Petit-Demouliere, B., Ilien, B., Simonin, F., 2013. Endogenous mammalian RF-amide peptides, including PrRP, kisspeptin and 26RFa, modulate nociception and morphine analgesia via NPFF receptors. Neuropharmacology 75, 164–171. Elshourbagy, N.A., Ames, R.S., Fitzgerald, L.R., Foley, J.J., Chambers, J.K., Szekeres, P.G., Evans, N.A., Schmidt, D.B., Buckley, P.T., Dytko, G.M., Murdock, P.R., Milligan, G., Groarke, D.A., Tan, K.B., Shabon, U., Nuthulaganti, P., Wang, D.Y., Wilson, S., Bergsma, D.J., Sarau, H.M., 2000. Receptor for the pain modulatory neuropeptides FF and AF is an orphan G protein-coupled receptor. J. Biol. Chem. 275, 25965–25971. Gaubert, G., Bertozzi, F., Kelly, N.M., Pawlas, J., Scully, A.L., Nash, N.R., Gardell, L.R., Lameh, J., Olsson, R., 2009. Discovery of selective nonpeptidergic neuropeptide FF2 receptor agonists. J. Med. Chem. 52, 6511–6514. Gouarderes, C., Zajac, J.M., 2007. Biochemical anti-opioid action of NPFF2 receptors in rat spinal cord. Neurosci. Res. 58, 91–94. Gouarderes, C., Jhamandas, K., Sutak, M., Zajac, J.M., 1996. Role of opioid receptors in the spinal antinociceptive effects of neuropeptide FF analogues. Br. J. Pharmacol. 117, 493–501. Gouarderes, C., Tafani, J.A., Zajac, J.M., 1998. Affinity of neuropeptide FF analogs to opioid receptors in the rat spinal cord. Peptides 19, 727–730. Gouarderes, C., Roumy, M., Advokat, C., Jhamandas, K., Zajac, J.M., 2000. Dual localization of neuropeptide FF receptors in the rat dorsal horn. Synapse 35, 45–52. Gouarderes, C., Quelven, I., Mollereau, C., Mazarguil, H., Rice, S.Q., Zajac, J.M., 2002. Quantitative autoradiographic distribution of NPFF1 neuropeptide FF receptor in the rat brain and comparison with NPFF2 receptor by using [125I]YVP and [(125I)]EYF as selective radioligands. Neuroscience 115, 349–361. Gouarderes, C., Faura, C.C., Zajac, J.M., 2004a. Rodent strain differences in the NPFF1 and NPFF2 receptor distribution and density in the central nervous system. Brain Res. 1014, 61–70. Gouarderes, C., Puget, A., Zajac, J.M., 2004b. Detailed distribution of neuropeptide FF receptors (NPFF1 and NPFF2) in the rat, mouse, octodon, rabbit, guinea pig, and marmoset monkey brains: a comparative autoradiographic study. Synapse 51, 249–269. Hinuma, S., Shintani, Y., Fukusumi, S., Iijima, N., Matsumoto, Y., Hosoya, M., Fujii, R., Watanabe, T., Kikuchi, K., Terao, Y., Yano, T., Yamamoto, T., Kawamata, Y., Habata, Y., Asada, M., Kitada, C., Kurokawa, T., Onda, H., Nishimura, O., Tanaka, M., Ibata, Y., Fujino, M., 2000. New neuropeptides containing carboxy-terminal RFamide and their receptor in mammals. Nat. Cell Biol. 2, 703–708. Hirsch, S., Corradini, L., Just, S., Arndt, K., Doods, H., 2013. The CGRP receptor antagonist BIBN4096BS peripherally alleviates inflammatory pain in rats. Pain 154, 700–707. Honor, P., Menning, P.M., Rogers, S.D., Nichols, M.L., Basbaum, A.I., Besson, J.M., Mantyh, P.W., 1999. Spinal substance P receptor expression and internalization in acute, short-term, and long-term inflammatory pain states. J. Neurosci. 19, 7670–7678. Jhamandas, J.H., Goncharuk, V., 2013. Role of neuropeptide FF in central cardiovascular and neuroendocrine regulation. Front Endocrinol (Lausanne) 4, 8. Kavaliers, M., Innes, D., 1992. Sex differences in the effects of neuropeptide FF and IgG from neuropeptide FF on morphine- and stress-induced analgesia. Peptides 13, 603–607.

Y.-T. Lin et al. / Experimental Neurology 291 (2017) 62–73 Kavaliers, M., Yang, H.Y., 1989. IgG from antiserum against endogenous mammalian FMRF-NH2-related peptides augments morphine- and stress-induced analgesia in mice. Peptides 10, 741–745. Kontinen, V.K., Kalso, E.A., 1995. Differential modulation of alpha 2-adrenergic and muopioid spinal antinociception by neuropeptide FF. Peptides 16, 973–977. Lameh, J., Bertozzi, F., Kelly, N., Jacobi, P.M., Nguyen, D., Bajpai, A., Gaubert, G., Olsson, R., Gardell, L.R., 2010. Neuropeptide FF receptors have opposing modulatory effects on nociception. J. Pharmacol. Exp. Ther. 334, 244–254. Lin, Y.T., Ro, L.S., Wang, H.L., Chen, J.C., 2011. Up-regulation of dorsal root ganglia BDNF and trkB receptor in inflammatory pain: an in vivo and in vitro study. J. Neuroinflammation 8, 126. Lin, Y.T., Kao, S.C., Day, Y.J., Chang, C.C., Chen, J.C., 2015. Altered nociception and morphine tolerance in neuropeptide FF receptor type 2 over-expressing mice. Eur. J. Pain 20 (6), 895–906 2016. Lin, Y.T., Liu, T.Y., Yang, C.Y., Yu, Y.L., Chen, T.C., Day, Y.J., Chang, C.C., Huang, G.J., Chen, J.C., 2016. Chronic activation of NPFFR2 stimulates the stress-related depressive behaviors through HPA axis modulation. Psychoneuroendocrinology 71, 73–85. Liu, Q., Guan, X.M., Martin, W.J., McDonald, T.P., Clements, M.K., Jiang, Q., Zeng, Z., Jacobson, M., Williams Jr., D.L., Yu, H., Bomford, D., Figueroa, D., Mallee, J., Wang, R., Evans, J., Gould, R., Austin, C.P., 2001. Identification and characterization of novel mammalian neuropeptide FF-like peptides that attenuate morphine-induced antinociception. J. Biol. Chem. 276, 36961–36969. Lombard, M.C., Weil-Fugazza, J., Ries, C., Allard, M., 1999. Unilateral joint inflammation induces bilateral and time-dependent changes in neuropeptide FF binding in the superficial dorsal horn of the rat spinal cord: implication of supraspinal descending systems. Brain Res. 816, 598–608. Malin, D.H., Henceroth, M.M., Izygon, J.J., Nghiem, D.M., Moon, W.D., Anderson, A.P., Madison, C.A., Goyarzu, P., Ma, J.N., Burstein, E.S., 2015. Reversal of morphine tolerance by a compound with NPFF receptor subtype-selective actions. Neurosci. Lett. 584, 141–145. Marchand, F., Perretti, M., McMahon, S.B., 2005. Role of the immune system in chronic pain. Nat. Rev. Neurosci. 6, 521–532. Matteoli, M., Haimann, C., Torri-Tarelli, F., Polak, J.M., Ceccarelli, B., De Camilli, P., 1988. Differential effect of alpha-latrotoxin on exocytosis from small synaptic vesicles and from large dense-core vesicles containing calcitonin gene-related peptide at the frog neuromuscular junction. Proc. Natl. Acad. Sci. U. S. A. 85, 7366–7370. Mouledous, L., Mollereau, C., Zajac, J.M., 2010. Opioid-modulating properties of the neuropeptide FF system. Biofactors 36, 423–429. Mouledous, L., Froment, C., Dauvillier, S., Burlet-Schiltz, O., Zajac, J.M., Mollereau, C., 2012. GRK2 protein-mediated transphosphorylation contributes to loss of function of muopioid receptors induced by neuropeptide FF (NPFF2) receptors. J. Biol. Chem. 287, 12736–12749. Mulderry, P.K., Ghatei, M.A., Bishop, A.E., Allen, Y.S., Polak, J.M., Bloom, S.R., 1985. Distribution and chromatographic characterisation of CGRP-like immunoreactivity in the brain and gut of the rat. Regul. Pept. 12, 133–143. Nystedt, J.M., Lemberg, K., Lintunen, M., Mustonen, K., Holma, R., Kontinen, V.K., Kalso, E., Panula, P., 2004. Pain- and morphine-associated transcriptional regulation of neuropeptide FF and the G-protein-coupled NPFF2 receptor gene. Neurobiol. Dis. 16, 254–262. Panula, P., Kalso, E., Nieminen, M., Kontinen, V.K., Brandt, A., Pertovaara, A., 1999. Neuropeptide FF and modulation of pain. Brain Res. 848, 191–196.

73

Porro, C.A., Baraldi, P., Pagnoni, G., Serafini, M., Facchin, P., Maieron, M., Nichelli, P., 2002. Does anticipation of pain affect cortical nociceptive systems? J. Neurosci. 22, 3206–3214. Rebeyrolles, S., Zajac, J.M., Roumy, M., 1996. Neuropeptide FF reverses the effect of muopioid on Ca2+ channels in rat spinal ganglion neurones. Neuroreport 7, 2979–2981. Ren, K., Dubner, R., 1999. Inflammatory models of pain and hyperalgesia. ILAR J. 40, 111–118. Roumy, M., Zajac, J.M., 1998. Neuropeptide FF, pain and analgesia. Eur. J. Pharmacol. 345, 1–11. Roumy, M., Zajac, J., 1999. Neuropeptide FF selectively attenuates the effects of nociceptin on acutely dissociated neurons of the rat dorsal raphe nucleus. Brain Res. 845, 208–214. Roumy, M., Lorenzo, C., Mazeres, S., Bouchet, S., Zajac, J.M., Mollereau, C., 2007. Physical association between neuropeptide FF and micro-opioid receptors as a possible molecular basis for anti-opioid activity. J. Biol. Chem. 282, 8332–8342. Roussin, A., Serre, F., Gouarderes, C., Mazarguil, H., Roumy, M., Mollereau, C., Zajac, J.M., 2005. Anti-analgesia of a selective NPFF2 agonist depends on opioid activity. Biochem. Biophys. Res. Commun. 336, 197–203. Russell, F.A., King, R., Smillie, S.J., Kodji, X., Brain, S.D., 2014. Calcitonin gene-related peptide: physiology and pathophysiology. Physiol. Rev. 94, 1099–1142. Seybold, V.S., 2009. The role of peptides in central sensitization. Handb. Exp. Pharmacol. 451–491. Shackman, A.J., Salomons, T.V., Slagter, H.A., Fox, A.S., Winter, J.J., Davidson, R.J., 2011. The integration of negative affect, pain and cognitive control in the cingulate cortex. Nat. Rev. Neurosci. 12, 154–167. Snijdelaar, D.G., Dirksen, R., Slappendel, R., Crul, B.J., 2000. Substance P. Eur. J. Pain 4, 121–135. Vera-Portocarrero, L.P., Westlund, K.N., 2004. Attenuation of nociception in a model of acute pancreatitis by an NK-1 antagonist. Pharmacol. Biochem. Behav. 77, 631–640. Vilim, F.S., Aarnisalo, A.A., Nieminen, M.L., Lintunen, M., Karlstedt, K., Kontinen, V.K., Kalso, E., States, B., Panula, P., Ziff, E., 1999. Gene for pain modulatory neuropeptide NPFF: induction in spinal cord by noxious stimuli. Mol. Pharmacol. 55, 804–811. Williams, K.A., Magnuson, M., Majeed, W., LaConte, S.M., Peltier, S.J., Hu, X., Keilholz, S.D., 2010. Comparison of alpha-chloralose, medetomidine and isoflurane anesthesia for functional connectivity mapping in the rat. Magn. Reson. Imaging 28, 995–1003. Winter, C.A., Risley, E.A., Nuss, G.W., 1962. Carrageenin-induced edema in hind paw of the rat as an assay for antiiflammatory drugs. Proc. Soc. Exp. Biol. Med. 111, 544–547. Yang, H.Y., Iadarola, M.J., 2003. Activation of spinal neuropeptide FF and the neuropeptide FF receptor 2 during inflammatory hyperalgesia in rats. Neuroscience 118, 179–187. Yang, H.Y., Iadarola, M.J., 2006. Modulatory roles of the NPFF system in pain mechanisms at the spinal level. Peptides 27, 943–952. Yang, H.Y., Fratta, W., Majane, E.A., Costa, E., 1985. Isolation, sequencing, synthesis, and pharmacological characterization of two brain neuropeptides that modulate the action of morphine. Proc. Natl. Acad. Sci. U. S. A. 82, 7757–7761. Yang, H.Y., Tao, T., Iadarola, M.J., 2008. Modulatory role of neuropeptide FF system in nociception and opiate analgesia. Neuropeptides 42, 1–18. Zeng, Z., McDonald, T.P., Wang, R., Liu, Q., Austin, C.P., 2003. Neuropeptide FF receptor 2 (NPFF2) is localized to pain-processing regions in the primate spinal cord and the lower level of the medulla oblongata. J. Chem. Neuroanat. 25, 269–278. Zubrzycka, M., Janecka, A., 2000. Substance P: transmitter of nociception (Minireview). Endocr. Regul. 34, 195–201.