Rab7—a novel redox target that modulates inflammatory pain processing

Abstract Chronic pain is accompanied by production of reactive oxygen species (ROS) in various cells that are important for nociceptive processing. Recent data indicate that ROS can trigger specific redox-dependent signaling processes, but the molecular targets of ROS signaling in the nociceptive system remain largely elusive. Here, we performed a proteome screen for pain-dependent redox regulation using an OxICAT approach, thereby identifying the small GTPase Rab7 as a redox-modified target during inflammatory pain in mice. Prevention of Rab7 oxidation by replacement of the redox-sensing thiols modulates its GTPase activity. Immunofluorescence studies revealed Rab7 expression to be enriched in central terminals of sensory neurons. Knockout mice lacking Rab7 in sensory neurons showed normal responses to noxious thermal and mechanical stimuli; however, their pain behavior during inflammatory pain and in response to ROS donors was reduced. The data suggest that redox-dependent changes in Rab7 activity modulate inflammatory pain sensitivity.


Introduction
Chronic pain is characterized by sensitization of the nociceptive system leading to increased responses to noxious stimuli (hyperalgesia), increased responses to normally innocuous stimuli (allodynia), and to spontaneous pain in the absence of any stimulus.As currently available treatments are only partially effective and associated with severe side effects, it is important to elucidate molecular mechanisms of nociceptive processing to develop novel strategies for chronic pain treatment. 1,30,31merging lines of evidence suggest that reactive oxygen species (ROS) such as superoxide (O 2 2 ) or hydrogen peroxide (H 2 O 2 ) are generated in nociceptive pathways and contribute to pain sensitization in a specific manner. 36,65A specific function of ROS in chronic pain processing is supported at the behavioral level by data from mice lacking the ROS-producing NADPH oxidases Nox1, Nox2, or Nox4.These mice showed reduced hypersensitivity in models of inflammatory or neuropathic pain. 26,27,32,35,38,43Moreover, administration of ROS scavengers and antioxidants reduced the pain behavior in various animal models, 39,47,62,70 whereas deficiency of antioxidant proteins enhanced the pain behavior. 2,34,69pecific redox signaling most often involves reversible oxidation of thiol residues within proteins thus changing structural and functional characteristics that may alter protein function. 37,64In general, several proteins, including transient receptor potential (TRP) ion channels, N-methyl-D-aspartate (NMDA) receptors, Ttype calcium channels, and GABA A receptors, have been found to sense ROS and change their activity in a redox-dependent manner. 13,40However, to date only few specific redox targets of ROS-dependent pain processing have been identified.Recent redox-proteome approaches make use of altered biotin switch techniques in combination with liquid chromatography-mass spectrometry (LC-MS)/mass spectrometry (MS) analysis which allows for screening of redox-modified proteins within tissues.To identify redox-regulated proteins in the nociceptive synapse of the spinal cord of mice, we took advantage of the OxICAT method. 42With this technique, reduced and oxidized thiols are differentially labeled using isotope-coded affinity tags (ICATs), which are subsequently detected and quantified by LC-MS/MS analysis.
One of the redox-regulated proteins we identified in our screen was the small Guanosine-59-triphosphate (GTP)ase Rab7 that plays a key role in late-endosomal trafficking, promoting fusion of endosomes, phagosomes, and autophagosomes with the lysosome. 75,76Rab7 activation is mediated by various regulators, including guanine-nucleotide exchange factors and other regulators such as homotypic fusion and protein sorting complexes, RILP (Rab7-interacting lysosomal protein) or FYCO1 (Fab1-YotB-Vac1p-EEA1 and coiled-coil domain containing 1) that facilitate Rab7 prenylation and membrane recruitment. 25,75Of interest, Rab7 shows a high homology between species and mutations of Rab7 or its interaction partners have been linked to various physiological disorders.For example, Rab7 mutations have been connected to Charcot-Marie-Tooth disease 2B (CMT2B), and defects of Rab5-and Rab7-mediated endosomal trafficking are related to Alzheimer disease, Down syndrome, and Parkinson disease. 5,8,10,12,24,56,73,74,76,77In this study, we characterized the expression of Rab7 in the nociceptive system and its putative pain-relevant functions.Our data suggest that Rab7 is a specific redox target that modulates inflammatory pain processing.

Animals
Tissue expression studies were performed in male C57BL/6N mice (Harlan).In addition, Rab7-deficient mice of either sex on a C57BL/6 background were used for behavioral studies.Generation of "floxed" Rab7 mice (Rab7 fl/fl ) has been described previously. 63To obtain conditional knockout mice lacking Rab7 in primary afferent neurons, Rab7 fl/fl mice were crossed with mice expressing the Cre recombinase under control of the Advillin promoter. 21Resulting heterozygous and homozygous conditional Rab7 knockouts are referred to as Adv-Rab 2/2 and Adv-Rab7 1/2 , respectively.Animals were housed on a 12/12 hours light/dark cycle with free access to water and food.All experiments were approved by the local Ethics Committee for Animal Research and adhered to the guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain.

Sample preparation
To quantitatively identify redox-modified cysteines in the spinal cord, the OxICAT protocol by Leichert et al. 42 was used with some adaptation to mouse tissue.The cleavable ICAT method development and cleavable ICAT reagent kits were obtained from AB SCIEX (Darmstadt, Germany).Fifteen microliters of a zymosan A suspension (5 mg/mL in 0.1 M phosphate-buffered saline [PBS], pH 7.4; Sigma-Aldrich, Schnelldorf, Germany) was injected into the plantar subcutaneous space of a mouse hind paw. 52Eight hours thereafter, lumbar spinal cords (L4-L5) of zymosan-injected and naive control animals were dissected out and immediately snap frozen in liquid nitrogen.Samples were weighed, homogenized in ice-cold 10% trichloroacetic acid (TCA), and a volume corresponding to 0.5 mg total tissue was used for further analysis.Samples were centrifuged for 30 minutes (13,000g, 4˚C), and TCA precipitates were consecutively washed in 10% and 5% ice-cold TCA.Resulting pellets were dissolved in denaturing alkylation buffer (8 M urea, 10 mM EDTA, 200 mM TrisHCl, pH 8.5) containing 0.5% sodium dodecyl sulfate (SDS).The contents of one light ICAT reagent vials were dissolved in 10 mL acetonitrile, and 5 mL were transferred to each sample.Samples were incubated for 75 minutes at 37˚C and 1300 rpm in the dark.Light ICAT reagent was removed by precipitation of proteins with ice-cold acetone over night at 220˚C.The samples were then washed twice with ice-cold acetone, dissolved in 50 mL denaturing alkylation buffer buffer, and reduced with 1 mL of reducing reagent (50 mM Tris(2-carboyethyl)phosphine [TCEP]) provided in the kit for 5 minutes at 37˚C and 1300 rpm.Heavy ICAT reagent was dissolved in acetonitrile and added to samples, which were incubated for 2 hours at 37˚C and 1300 rpm in the dark.Samples were digested with trypsin for 16 hours at 37˚C.Peptides were acidified by adding 20 mL pure formic acid and purified with solid phase extraction cartridges (Empore 4115(SD); Sigma Aldrich, St. Paul, MA).Enrichment of ICAT-labeled peptides on avidin columns and cleavage of biotin moiety from the isotope coded tag were performed as described in the manufacturers' manual.

LC-MS/MS analysis
LC/MS was performed on a Thermo Scientific Orbitrap XL mass spectrometer linked to an Agilent1200 nano-high-performance liquid chromatography.Peptides were resolved in 4% acetonitrile and 0.5% formic acid and loaded on a C18 reversed-phase precolumn (Zorbax 300SB-C18; Agilent Technologies, Palo Alto, CA) followed by separation on an in-house packed 3 mm Reprosil C18 resin (Dr.Maisch GmbH, Ammerbuch, Germany) picotip emitter tip (diameter 75 mm, length 10 cm; New Objectives, Woburn, MA) using a gradient from 5% acetonitrile, 0.1% formic acid to 50% acetonitrile, 0.1% formic acid for 90 minutes with a flow rate of 200 nL/min.MS data were recorded by data-dependent Top10 acquisition selecting the 10 most abundant precursor ions in positive mode for fragmentation using dynamic exclusion of 3 minutes.Only highly charged ions (21) were selected for MS/MS scans in the linear ion trap by CID at 35% collision energy.Lock mass option for m/z 445.120,025 was enabled to ensure high mass accuracy during many following runs. 57

Data analysis
MS data were searched using Mascot Server (version 2.2) against a database containing species-specific (Mus musculus) protein sequences downloaded from www.uniprot.org.The Mascot search settings were as follows: maximum missed cleavages 1, precursor mass tolerance 10 ppm, fragment ion tolerance 0.8 Da, and optional modifications allowed on methionine (oxidation) and cysteine.Only peptides with individual ion scores indicating identity (P , 0.05) were considered significant.Quantification of heavy/light ICAT ratios of peptides was performed by Proteome Discoverer 1.2.Peptides containing cysteines were selected, and oxidation status (percentage) was calculated and further statistically analyzed.

Cell culture
For the redox-switch assay, human neuroblastoma cells (SH-SY5Y) were used.Cells were maintained in a 1:1 mix of minimal essential medium (ATCC, Wesel, Germany) and F-12 Nutrient Mix (Gibco, Carlsbad, CA) containing 15% fetal calf serum, 1% glutamine, 1% nonessential amino acid mix, and 1% penicillinstreptomycin at 5% CO 2 and 37˚C, and passaged once a week.Cells were cultured for 3 days before stimulation with 100 mM H 2 O 2 for 5 minutes. 45n addition, murine neuroblastoma cells (Neuro-2a) were used for transfection of Rab7 plasmids.Cells were cultivated in advanced DMEM containing 10% fetal calf serum (FCS), 1% glutamine, and 1% penicillin-streptomycin at 5% CO 2 and 37˚C and passaged twice a week.Transfection of cells was performed as described below.

Switch assay
Detection of oxidized thiols in the Rab7 protein was performed according to previous published methods. 4,15,16,78SH-SY5Y cells were stimulated with 100 mM H 2 O 2 as described above and homogenized in homogenization buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 25 mM NEM, 2% Triton-X, protease-inhibitor mix) and incubated for 5 minutes at room temperature (RT), followed by incubation for 5 minutes on ice.Cells were centrifuged at 16,000g and 4˚C for 20 minutes, and the supernatant was precipitated with ice-cold acetone.Pellets were solubilized in solubilization buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 2% Triton-X, protease-inhibitor mix) and reduced with TCEP (final concentration 2 mM) for 1 hour at 37˚C.Cy5-labeled maleimide was added, and samples and were incubated for another hour at 37˚C.Samples were then precipitated with ice-cold acetone and solubilized again in solubilization buffer.Rab7 protein was immunoprecipitated with rabbit anti-Rab7 (1:500; Sigma-Aldrich # R4779) and protein A/G agarose beads (Santa Cruz Biotechnology, Dallas, TX).Immunoprecipitated proteins were separated with SDS polyacrylamide gel electrophoresis and blotted onto a nitrocellulose membrane.Cy5-labeled (originally oxidized) proteins were detected using an Odyssey Infrared Imaging System (LI-COR Bioscience, Bad Homburg, Germany).Afterwards, membranes were blocked in blocking buffer (Odyssey blocking buffer, diluted 1:1 in PBS; LI-COR Bioscience) for 1 hour at RT and then incubated with rabbit anti-Rab7 (1:500; Cell Signaling Technology, Cambridge, UK, # 93679) dissolved in blocking buffer containing 0.1% Tween-20 overnight at 4˚C, followed by incubation with a secondary antibody for 1 hour at RT. Detection of Rab7 total protein was again performed using the Odyssey Infrared Imaging System, and quantification of band densities was done using NIH ImageJ software.

Cloning of the Rab7 "redox-dead" plasmid
Cloning of the a redox-insensitive Rab7 construct, further termed as "redox-dead," was performed using the Q5-Site-Directed Mutagenesis Kit (New England Biolabs, Frankfurt, Germany) according to the manufacturer's instructions.Primers were designed using the NEBase Exchanger v1.2.1 (Q5SDM_fwd 59agg tgc aga cag cag tgt tct ggt att tg -39, Q5SDM_rev 59-ctg tag aag gcc aca ccg -39) to exchange nucleotides coding for the 2 cysteines in position 83 and 84 of Rab7 (NCBI accession Number CAJ18560.1)to serine.Plasmid containing green fluorescent protein (GFP)-rab7 wildtype (WT) (clone 12605; Addgene Inc, Cambridge, MA) was used as template DNA.PCR amplification, ligation, and transformation were performed according to the manufacturer's protocol.NEB 5-alpha competent E.coli containing the Rab7 redox-dead plasmid were plated on kanamycincontaining agar plates.Plasmid isolation of selected clones was performed using the QIAprep Spin Miniprep Kit.Sequence verification was performed by LGC Genomics (Berlin, Germany).
Transfection of plasmids was performed using FuGENEHD Transfection Reagent (Promega, Mannheim, Germany) according to the manufacturer's protocol.Neuro-2a cells were seeded into 12-well or 6-cm dishes at a density of 4 3 10 3 and 4 3 10 5 , respectively.Transfection was performed immediately after seeding using a ratio of 5:2 of FuGENE transfection reagent and plasmid DNA.

Rab7 activity assay
Activity status of Rab7 constructs was assessed using an Rab7 activation assay kit (NewEast Bioscience, Malvern, PA).The assay was performed according to the manufacturer's instructions.Briefly, Neuro-2a cells were transfected with Rab7-WT or Rab7 redox-dead constructs and cultured for 2 days.Cells were then lysed in assay buffer, centrifuged, and supernatants were loaded either with GTP or Guanosine-59-diphosphate (GDP) to stimulate Rab7.Active/GTP-bound Rab7 was then immunoprecipitated and detected by immunoblot.Values were calculated to the respective GDP loaded control.

Real-time reverse transcription PCR
Total RNA from the lumbar spinal cord (L4-L5), lumbar DRG (dorsal root ganglia) (L4-L5), and the sciatic nerve was extracted under RNase-free conditions using TRI-reagent (Sigma-Aldrich) or the RNAqueous Micro Kit (Ambion/Life Technologies, Darmstadt, Germany).Reverse transcription was performed using a First Strand cDNA Synthesis Kit (Thermo Scientific, Waltham, MA) according to the manufacturer's instructions.Quantitative real-time RT-PCR was performed with an ABI Prism 7500 Sequence Detection System (Applied Biosystems/Life Technologies) using gene-specific primer pairs for murine Rab7 (fwd 59-gga aga aag tgt tgc tga agg -39, rev 59-tgt ggc ttt gta ctg gtt act g-39) and GAPDH (fwd 59-caa tgt gtc cgt gga tct -39, rev 59gtc ctc agt gta gcc caa gat g-39).Reactions were performed in duplicate or triplicate by incubating for 10 minutes at 95˚C, followed by 40 cycles of 15 seconds at 95˚C and 1 minute at 60˚C, including water controls to ensure specificity.Relative gene expression levels were calculated using the comparative 2 2DDCt method and normalized to GAPDH.

In situ hybridization
Mice were killed by CO 2 and intracardially perfused with 4% formaldehyde (PFA) in PBS for 5 minutes.Dorsal root ganglia were dissected and postfixed in the same fixative for 10 minutes.Tissue was then incubated in 20% sucrose in PBS overnight, embedded in Tissue Tek mounting media (Sakura Finetek, Alphen aan den Rijn, Netherlends), and cryostat sections were cut at a thickness of 16 mm.In situ hybridization (ISH) was performed using an Affymetrix QuantiGene ViewRNA Tissue Assay in which target mRNA signals appear as puncta in microscopy. 58,72Experiments were performed according to the instructions of the manufacture and included extensive washing steps in PBS.Briefly, sections were fixed in 4% PFA overnight, dehydrated for 10 minutes in a 50%, 70%, and 100% ethanol series, dried for 30 minutes at 60˚C in a ThermoBrite slide processing system (Abbott Molecular, Des Plaines, IL), and digested with the Protease QF Mix for 20 minutes.Then sections were fixed again in 4% PFA for 5 minutes and hybridized overnight at 40˚C (ThermoBrite) with an approximately 600 nucleotides spanning type 1 probe set designed by Affymetrix to mouse Rab7 (NCBI accession number NM_009005, catalog # VB1-20408).In control experiments, a scramble probe (catalog # VF1-17155) was used.Sections were then consecutively incubated with PreAmplifier Mix QT, Amplifier Mix QT, an alkaline phosphataselabeled probe against the Amplifier, AP Enhancer Solution, and Fast Red Substrate (Cy3 fluorescence).Finally, sections were coverslipped with Fluoromount-G (Southern Biotech) or further processed for subsequent immunohistochemistry.
In ISH experiments combined with immunohistochemistry, hybridization of Rab7 was performed as described above.Sections were then blocked in 10% NGS and 3% bovine serum albumin in PBS-Triton-X 100 for 1 hour and incubated with primary antibodies overnight.The following antibodies were used: rabbit anti-NF200 (Sigma-Aldrich, # N4142) and mouse anti-peripherin (Chemicon, # MAB1527).After rinsing in PBS, slides were incubated with secondary antibodies conjugated with Alexa Fluor 488 or Cy5 (Invitrogen/Life Technologies) for 2 hours at RT. Lipofuscin-like autofluorescence was reduced as described above, and slides were coverslipped with VECTASHIELD HardSet Antifade Mounting Medium with DAPI (Vector Laboratories, Burlingame, VT, # H-1500).Images were taken and adjusted as described above.

Cell counting
To analyze the distribution of Rab7 mRNA in DRG neurons, only Rab7-positive profiles with a clear nucleus were analyzed.The percentage of all DRG neurons (visualized by the autofluorescence of DRG neurons) positive for Rab7 and of Rab7-positive DRG neurons expressing peripherin or NF200 was calculated.In total, .2000DRG neurons of 6 animals (at least 150 cells per animal) were counted.
To analyze DRG neuron subpopulations of Adv-Rab7 1/2 mice, the percentage of neurons positive for marker (IB4, NF200) was calculated from 4 animals per genotype and 2 sections per animal (.2000 cells per genotype).Only DRG neurons showing staining clearly above background were included.Sections were chosen randomly with at least 70 mm apart.

Western blot
Lumbar (L4-L5) spinal cords, the dorsal part of lumbar spinal cords, lumbar DRG, and sciatic nerves were rapidly dissected, frozen in liquid nitrogen, and stored at 280˚C until use.Samples were homogenized in PhosphoSafe buffer (Novagen/Merck Millipore, Darmstadt, Germany) combined with a protease inhibitor cocktail (Complete Mini; Roche Diagnostics, Mannheim, Germany) and centrifuged at 14,000g for 1 hour.Extracted proteins (20 mg per lane) were separated by SDS polyacrylamide gel electrophoresis and blotted onto a nitrocellulose membrane.Membranes were blocked in blocking buffer (Odyssey blocking buffer, diluted 1:1 in PBS; LI-COR Bioscience) for 1 hour at RT and then incubated with rabbit anti-Rab7 (1:200; Cell Signaling # 93679, Danvers, MA) or mouse anti-GAPDH (1:2000; Ambion # AM4300) dissolved in blocking buffer containing 0.1% Tween-20 overnight at 4˚C.After incubation with secondary antibodies for 1 hour at RT, proteins were detected using an Odyssey Infrared Imaging System (LI-COR Bioscience).Quantification of band densities was performed using NIH ImageJ software.

Behavioral tests
Behavioral studies were performed by a blinded observer.Experiments were performed using littermate mice of either sex at the age of 7 to 12 weeks.

Rotarod test
A rotarod treadmill for mice (Ugo Basile, Varese, Italy) was used to assess motor coordination, running at a constant rotating speed of 12 rpm.After several training sessions, fall-off latencies from 3 to 5 tests were averaged.The cut-off time was set at 120 seconds.

Hot-plate and cold-plate test
Mice were placed on a heated or cooled metal plate surrounded by a plexiglas cylinder (Hot/Cold Plate; Ugo Basile, Monvalle, Italy).For the hot-plate test, the latency time between placing and the first nociceptive reaction (shaking or licking) of a hind paw was recorded.Experiments were performed at temperatures of 50, 52, and 54˚C with cut-off times of 60, 40 and 20 seconds, respectively, to prevent tissue damage.In the cold-plate test, the metal plate was cooled down at 0˚C, and the number of reactions over a period of 5 minutes was counted.Hind paw withdrawal, hind paw licking and jumping were considered as a reaction. 41

Dynamic plantar test
Mechanical sensitivity of the hind paw was analyzed using a dynamic plantar aesthesiometer (Ugo Basile).The device pushes a thin probe (0.5 mm diameter) with increasing force through a wire-gated floor against the plantar surface of the paw from beneath.When the animal withdraws the paw, the device automatically stops and records the latency time, which is the time from when the probe touched the paw to when the animal withdrew the paw.The force was increased from 0 to 5 g within 10 seconds (0.5 g/s ramp) and then held at 5 g for an additional 10 seconds.Paw withdrawal latencies were calculated as the mean of 4 to 6 measurements with at least 20 seconds in between. 35,46,6712.4.Zymosan-induced hyperalgesia Fifteen microliters of a zymosan A suspension (5 mg/mL in 0.1 M PBS, pH 7.4; Sigma-Aldrich) was injected into the plantar subcutaneous surface of a hind paw (Meller and Gebhart,  1997).Mechanical sensitivity of the affected hind paw was measured using the dynamic plantar test as described above.Thermal hypersensitivity was assessed using a plantar test (Hargreaves Apparatus; Ugo Basile).Infrared light beam intensity was set to 25, which resulted in paw withdrawal latencies of 8 to 10 seconds for baseline measurements.

Reactive oxygen species-induced pain behavior
The ROS donor, tert-butyl hydroperoxide (TBHP; 2.2 mmol), dissolved in 20 mL saline, was injected subcutaneously into the dorsal surface of a hind paw and the time spent licking the TBHP-injected paw was measured over a period of 10 minutes.The TBHP dose was selected on the basis of previous studies. 34,47ly 2017 • Volume 158 • Number 7 www.painjournalonline.com1357 Copyright Ó 2017 by the International Association for the Study of Pain.Unauthorized reproduction of this article is prohibited.

Identification of reactive oxygen species targets in chronic pain processing
To identify pain-relevant ROS targets, we performed the wellcharacterized in vivo model of zymosan-evoked hind paw inflammation in mice. 22,33,48,52Eight hours after the zymosan injection (ie, at a time point when mechanical hyperalgesia was pronounced, as indicated by a drop in paw withdrawal latencies, Fig. 1A), the lumbar spinal cord was removed and analyzed using the OxICAT thiol-trapping technique, 42 which allows for screening redox-modulated thiol groups in proteins.As shown in Table 1, we identified 10 proteins that were significantly redox modulated in zymosan-injected mice as compared to naive control mice ($2 fold changes in the redox-state; P , 0.05; n 5 3-4 animals per group).Of interest, the protein found to be most strongly oxidized after zymosan injection was the small GTPase Rab7 (Fig. 1B).Moreover, the sequence detected by the OxICAT screen is highly conserved among species (Fig. 1C).Previous studies suggest a contribution of Rab7 to pain processing; 55,66,77 however, its possible function as a pain-relevant ROS target has not been investigated yet.We therefore focused on Rab7 for the remainder of our study.

Rab7 is a reactive oxygen species target in vitro
To confirm that Rab7 may serve as a target for ROS-based signaling, we performed a modified switch assay in which oxidized thiols are labeled with maleimide-Cy5, and the protein of interest is then enriched by immunoprecipitation. 15,16,78 As shown in Figure 2, stimulation of SH-SY5Y neuroblastoma cells with 100 mM H 2 O 2 resulted in increased oxidation of Rab7 as indicated by increased Rab7-Cy5 labeling.These in vitro data support our observation that Rab7 is sensitive to ROS-dependent oxidation.

Rab7 expression in the spinal cord and dorsal root ganglia
We then investigated the distribution of Rab7 in the murine spinal cord using immunostaining.We found Rab7 expression to be enriched in the superficial laminae of the dorsal horn and in distinct neurons of the ventral horn (Fig. 3A).Double-labeling experiments revealed Rab7 expression in calcitonin gene-related peptide-and isolectin B4 (IB4)-positive fibers (Fig. 3B, C), which are markers for peptidergic and non-peptidergic unmyelinated nociceptors, respectively.Co-localization was also detected for Rab7 and NF200 (Fig. 3D), a marker of large myelinated primary afferents.Specificity of anti-Rab7 was confirmed by the reduced immunoreactivity in the spinal cord of mice lacking Rab7 in sensory neurons (Adv-Rab7 2/2 , Fig. 3E).These data suggest that Rab7 is enriched in central sensory neuron terminals and in motoneurons of the spinal cord of naive mice.
We next analyzed the Rab7 localization in DRG.As immunostaining experiments using various protocols failed to detect specific Rab7 immunoreactivity in DRG, we performed fluorescent ISH of Rab7 mRNA.We found Rab7 to be expressed in the majority of DRG neurons (89.2% 6 4.3%; Fig. 4A).As expected, no hybridization signals were detected using a scramble control probe (Fig. 4B).To further characterize the localization of Rab7 in DRG neuron subpopulations, we combined ISH of Rab7 mRNA with immunostaining for the marker peripherin and NF200, which stain unmyelinated and myelinated DRG neurons, respectively.We found that  Copyright Ó 2017 by the International Association for the Study of Pain.Unauthorized reproduction of this article is prohibited.
We then performed Western blot experiments to analyze whether Rab7 expression in DRG and the dorsal horn of the spinal cord is altered during zymosan-induced hind paw inflammation.Of interest, we found Rab7 expression to be significantly increased in ipsilateral DRG 4 and 8 hours after the zymosan injection (Fig. 5A), whereas Rab7 expression was significantly reduced in the ipsilateral dorsal horn of the spinal cord (Fig. 5B).Quantitative real-time RT-PCR analyses revealed that Rab7 mRNA was not upregulated in ipsilateral DRG 6 hours after the zymosan injection (relative Rab7 mRNA expression: naive mice, 1.00 6 0.06; zymosan-injected mice, 1.15 6 0.17; P 5 0.48; n 5 3).Hence, these data point to a zymosan-induced translocation of Rab7 protein to the somata of DRG neurons.This assumption is supported by an earlier study in which Rab7 translocation to perinuclear lysosome compartments of DRG neurons was observed during diabetic neuropathy in rats. 55Altogether, these data point to a contribution of Rab7 to pain processing.

Rab7 knockdown mice display normal basal sensitivity
To characterize the function of Rab7 in sensory neurons in vivo, we crossed mice carrying the floxed exon 1 of the Rab7 gene 63 with Advillin-Cre mice, 21,79 to obtain mice lacking Rab7 specifically in primary afferent neurons.Unexpectedly, we only obtained a relatively small proportion (8.9%) of homozygous mutant mice (Adv-Rab7 2/2 ) in the offspring suggesting that homozygous mutant mice already die within the late prenatal period.This observation is in accordance with a previous study in Drosophila in which a complete loss of Rab7 severely impaired pupal development and with the observation that mutations in the Rab7 gene caused defects in sensory neuron axon growth, branching, and path finding in a zebrafish model. 7,59We therefore used heterozygous mutant mice (Adv-Cre; Rab7 1/fl ; referred to as Adv-Rab7 1/2 ) in our behavioral studies.
Quantitative real-time RT-PCR analyses revealed a reduced Rab7 mRNA expression in DRG of Adv-Rab7 1/2 mice  compared with control mice, whereas mRNA expression levels in the spinal cord, the kidney, and the heart were similar between groups (Fig. 6A).At the protein level, Rab7 expression was reduced in tissue lysates of the spinal cord, DRG, and the sciatic nerve of Adv-Rab7 1/2 mice (Fig. 6B).Moreover, the proportion of DRG neurons positive for IB4 and neurofilament 200 (NF200) was similar between genotypes (Fig. 6C).Hence, Rab7 expression is moderately reduced at sites of Advillin-Cre-mediated recombination in Adv-Rab7 1/2 mice.

Inflammatory and reactive oxygen species-induced pain behavior is reduced in Rab7 knockdown mice
We next assessed the behavior of Adv-Rab7 1/2 and control mice in the zymosan model of inflammatory pain, in which Rab7 oxidation was detected (Fig. 1).Injection of zymosan into a hind paw resulted in mechanical hypersensitivity in both genotypes.Notably, the extent of hypersensitivity was significantly lower in Adv-Rab7 1/2 mice as compared to control littermates (Fig. 8A).In contrast to mechanical hypersensitivity, thermal hypersensitivity was not significantly affected by the Rab7 knockdown (Fig. 8B).These data suggest that Rab7 functionally contributes to mechanical, but not thermal, hypersensitivity during the processing of inflammatory pain.
We then investigated the pain behavior in response to ROS delivery.For this purpose, we injected the ROS donor TBHP (2.2 mmol) into a hind paw and observed the induced paw licking behavior over a period of 10 minutes.As shown in Figure 8C, Adv-Rab7 1/2 mice demonstrated significantly less paw licking compared with control littermates, suggesting that Rab7 is a target of ROS-induced pain sensitization in vivo.Altogether, these behavioral data in combination with our OxICAT screen indicate functional contribution of Rab7 oxidation to inflammatory pain processing.

Redox modifications affect Rab7 activity in vitro
To assess the functional implications of Rab7 oxidation, we transfected Neuro-2a neuroblastoma cells with Rab7-WT or an Rab7 redox-dead construct, in which the redox-sensitive thiol residues of Rab7 were each replaced by serine (Fig. 9A).In both constructs, Rab7 was tagged to GFP resulting in an enhanced molecular weight (50 kDa) compared with native Rab7 (23 kDa) as detected by Western Blot (Fig. 9B), allowing us to distinguish between native and transfected Rab7.In the transfected cells, we determined Rab7 activity using an Rab7 activation assay kit.Surprisingly, we found that after GTP loading, Rab7 activity was higher in cells transfected with Rab7 redox-dead cells (Fig. 9C).We further assessed subcellular localization of native and mutant Rab7 by immunostaining, but there was no obvious difference (Fig. 9D).Hence, oxidation of Rab7 suppresses its GTP-binding activity suggesting that the observed oxidation in the spinal cord in the context of inflammatory pain processing is associated with reduced GTPase activity.

Discussion
In this study, we show that redox-mediated posttranslational protein modifications contribute to inflammatory pain hypersensitivity.In an OxICAT screen, we identified Rab7, which is highly expressed in sensory neurons and their central terminals in the superficial laminae of the spinal cord, as an oxidized target during pain processing.We further show that the activity of Rab7 was altered in vitro in a redox-dependent manner and that heterozygous Rab7 deficiency reduced inflammatory and ROS-induced pain behavior.
In the last few years, our understanding of redox signaling has increased dramatically.Although ROS were initially considered as harmful agents that attack proteins, lipids, or nucleic acids, it is now clear that they serve as signaling molecules that modify protein activity at a posttranscriptional level.Oxidative modifications may alter the functions of the respective targeted cysteine residues, and this may result in changes in enzyme activity, binding properties, stability, or localization.Redox-dependent protein modifications include the formation of intra-or inter-protein disulfides, S-sulfenation, S-sulfination, S-sulfonation, S-nitrosylation, S-sulfhydration, or S-sulfenamidation. 29,44,64Advances in biochemical and molecular biology methods have allowed the identification and quantification of redox-modified proteins.These methods often make use of different labeling and trapping of the redox status of protein residues that stabilize the relatively labile oxidationmodified cysteines with stable alkylated tags which can then be detected by electrophoresis, enzyme-linked immunosorbent assay, or mass spectrometry (MS) analysis. 23,64Although the OxICAT-labeling technique we used here revealed several redox-modified proteins, the method has its limitations.For example, higher oxidation states of the thiolate ion such as Ssulfination or S-sulfonation are not detected by OxICAT.Furthermore, the method preferentially detects abundant proteins.One could argue that the time point we chose (ie, 8 hours after zymosan injection) was associated with an increased protein expression that resulted in false-positive OxICAT hits.However, we did observe significant zymosaninduced Rab7 downregulation in the spinal cord (Fig. 5B), suggesting that the changes we detected in the OxICAT screen were due to oxidation and not upregulation of Rab7 protein.We only analyzed proteins which were reliably detected in at least 3 of 4 samples.Thus, less abundant proteins, which were not detected in all samples, might also be involved in redox signaling in the dorsal horn.
Several studies suggest a function of ROS as signaling molecules in various pain states.For example, administration of ROS scavengers and antioxidants reduced nociceptive responses in inflammatory pain models, 19,39,70 and impaired ROS production in knockout mice was associated with reduced pain hypersensitivity. 26,27,32,35,38,43However, so far, only a few redox targets have been found at sites of nociceptive signal transduction.The small GTPase Rab7, which we identified as a redox target in the nociceptive system, is an essential protein, regulating endosomal maturation and controlling various steps in membrane trafficking. 75,76Mutations in the Rab7 gene that result in increased Rab7 activity are associated with the Charcot-Marie-Tooth type 2B (CMT2B) disease, a peripheral neuropathy that is characterized by sensory loss, near-normal nerve conduction velocities, foot deformations, and a high frequency of ulcers and infections. 3,8,17In sensory neurons, Rab7 plays an essential role in axonal retrograde transport, acting in a sequential manner together with Rab5, thus controlling trafficking and neurite outgrowth. 6,9,11,77Of interest, an earlier study showed that Rab7 co-localizes with the ROS-producing enzyme, tartrate-resistant acid phosphatase in alveolar macrophages. 60oreover, infections with the human hepatitis virus strongly depend on Rab7-dependent endosomal transport which involves alterations in the redox state of the disulfide-bound envelope proteins of the virus. 49These studies imply that  redox-dependent modifications of Rab7 may also arise in other tissues and cellular systems.Here, we show that a redoxinsensitive mutant of Rab7 displays enhanced GTP binding, suggesting that oxidation of Rab7 terminates GTP-dependent activity of the protein.Of interest, redox-dependent processes have been recently considered as major regulators of autophagy. 14As Rab7 is a key regulator of autophagy, 25 it remains to be elucidated whether redox-dependent modifications of Rab7 affect autophagosomal signaling in general.
We found that Rab7-knockdown mice display reduced inflammatory and ROS-induced pain behavior.In the inflammatory pain model, behavioral responses to mechanical but not thermal stimuli were significantly affected by the Rab7 knockdown, suggesting that Rab7 predominantly contributes to mechanical hypersensitivity during inflammatory pain processing.The selective role of Rab7 in zymosan-induced mechanical hyperalgesia may be explained by distinct contribution to nociceptive pathways such as TRPV1-and TRPV4-dependent pain signaling, which modulate thermal and mechanical hypersensitivity, respectively.Although TRPV1 predominantly seems to modulate thermal hyperalgesia, TRPV4 is involved in mechanical hypersensitivity after zymosaninduced hind paw inflammation. 71Of interest, both Rab7 and TRPV4 have been associated with Charcot-Marie-Tooth disease. 17,50Moreover, Rab7 translocated from the central terminals to the somata of sensory neurons after zymosan injection into a hind paw, further supporting our observation that Rab7 affects inflammatory pain signaling.
Previous studies point towards interactions of Rab7 with different players in pain signaling.For example, during diabetic neuropathy in rats, Rab7 co-localizes with m-opioid receptors in perinuclear lysosome compartments, thereby promoting lysosomal targeting of m-opioid receptors and reducing opioid responsiveness. 55Furthermore, Rab7 controls endosomal trafficking of the nerve growth factor receptor TrkA, a key regulator of hyperalgesia. 51,53,66,77Rab7 also interacts with calcium-/calmodulin-dependent protein kinase 4, which has been implicated in nicotine-induced pain processing. 20,28ence, Rab7 as a key regulator of endosomal traffic might control various pain-relevant signaling mechanisms.Altogether, these findings are consistent with the idea that redox modulations of distinct proteins such as Rab7 make an essential contribution to pain processing.
1356 W. Kallenborn-Gerhardt et al. • 158 (2017) 1354-1365 PAIN ® Copyright Ó 2017 by the International Association for the Study of Pain.Unauthorized reproduction of this article is prohibited.

Figure 1 .
Figure 1.An OxICAT screen detects Rab7 redox modifications in the spinal cord after zymosan injection into a hind paw.(A) Mechanical hyperalgesia 8 hours after the zymosan injection, as indicated by a drop in withdrawal latencies on mechanical hind paw stimulation using a Dynamic Plantar Aesthesiometer (n 5 4, *P , 0.05).This time point was chosen for OxICAT analysis of spinal cord tissue.Identified proteins are presented in Table 1.(B) Positions of redox modifications in the Rab7 peptide.The structure of GDP-bound (blue) Rab7 (green) from rat [PDB database: 1vg1 61 ] was used to visualize redox-modified cysteines (magenta).(C) Alignment of Rab7 protein sequence of amino acids 71 to 100 in different species.The gray marked sequence (with redox-sensitive cysteines in positions 83 and 84) was detected by the OxICAT screen.This Rab7 protein sequence is highly conserved among species.

Figure 3 .
Figure 3. Localization of Rab7 in the spinal cord.(A) Rab7 expression is enriched in the superficial layer of the dorsal horn and in some ventral horn neurons.(B-D) Rab7 co-localizes with CGRP-(B) and IB4-(C) positive unmyelinated nociceptors.There is also some co-localization with NF200 (D), a marker for large myelinated neurons.(E) Rab7 expression in the dorsal horn is reduced in Adv-Rab7 2/2 mice, in which Rab7 is selectively knocked out in sensory neurons (scale bars: A 5 200 mm, B-D 5 25 mm, E 5 100 mm).CGRP, calcitonin gene-related peptide.

Figure 5 .
Figure 5. Rab7 expression after zymosan-induced hind paw inflammation.(A) Western blot analyses detected increased Rab7 protein levels in DRG 4 hours and 8 hours after zymosan injection into a hind paw.(B) In contrast, Rab7 protein levels in the ipsilateral dorsal horn of the spinal cord were significantly decreased 8 hours after the zymosan injection.n 5 3 to 4, *P , 0.05.DRG, dorsal root ganglia.

Figure 7 .
Figure 7. Acute nociceptive pain is normal in Rab7-knockdown mice.Adv-Rab7 1/2 mice displayed normal sensitivity to acute thermal stimuli on a 50 to 54˚C hot plate (A; n 5 12-13) and a 0˚C cold plate (B; n 5 8-10), and after mechanical stimulation of the hind paw using a Dynamic Plantar Aesthesiometer (C; n 5 16).
For paired comparisons and multiple comparisons, the unpaired, 2-tailed Student t test or 2way analysis of variance was used, the latter followed by t tests using a Bonferroni correction.Rotarod fall-off latencies were analyzed with the Mann-Whitney U test and are presented as medians and interquartile ranges.Other data are expressed as the mean 6 SEM.For all tests, a probability value P , 0.05 was accepted as statistically significant.Number of mice and replicates are given in the figure legends.

Table 1
Redox-modulated proteins identified in the OxICAT screen.