Satellite glial GPR37L1 and its ligand maresin 1 regulate potassium channel signaling and pain homeostasis

G protein–coupled receptor 37-like 1 (GPR37L1) is an orphan GPCR with largely unknown functions. Here, we report that Gpr37l1/GRP37L1 ranks among the most highly expressed GPCR transcripts in mouse and human dorsal root ganglia (DRGs) and is selectively expressed in satellite glial cells (SGCs). Peripheral neuropathy induced by streptozotoxin (STZ) and paclitaxel (PTX) led to reduced GPR37L1 expression on the plasma membrane in mouse and human DRGs. Transgenic mice with Gpr37l1 deficiency exhibited impaired resolution of neuropathic pain symptoms following PTX- and STZ-induced pain, whereas overexpression of Gpr37l1 in mouse DRGs reversed pain. GPR37L1 is coexpressed with potassium channels, including KCNJ10 (Kir4.1) in mouse SGCs and both KCNJ3 (Kir3.1) and KCNJ10 in human SGCs. GPR37L1 regulates the surface expression and function of the potassium channels. Notably, the proresolving lipid mediator maresin 1 (MaR1) serves as a ligand of GPR37L1 and enhances KCNJ10- or KCNJ3-mediated potassium influx in SGCs through GPR37L1. Chemotherapy suppressed KCNJ10 expression and function in SGCs, which MaR1 rescued through GPR37L1. Finally, genetic analysis revealed that the GPR37L1-E296K variant increased chronic pain risk by destabilizing the protein and impairing the protein’s function. Thus, GPR37L1 in SGCs offers a therapeutic target for the protection of neuropathy and chronic pain.

protocol No.CM130).After reaching 70% confluency by co-transfection with 1 µg of pmaxGFP control cDNA (Lonza), the transfected cells were cultured for an additional 48 h before use.

Human SGC/neuron-rich culture
Human DRG cultures were prepared as previously reported (1).DRGs were digested at 37°C in a humidified CO2 incubator for 120 min with collagenase Type II (Worthington, 290 units/mg, 12 mg/ml final concentration) and Dispase II (Roche, 1 unit/mg, 20 mg/mL) in PBS with 10 mM HEPES, pH adjusted to 7.4 with NaOH.DRGs were mechanically dissociated using fire-polished pipettes, filtered through a 100-m nylon mesh, and centrifuged (500 g for 5 min).The pellet was resuspended, plated on 0.5 mg/mL poly-D-lysine-coated glass coverslips, and grown in Neurobasal medium supplemented with 10% FBS, 2% B-27 supplement, and 1% penicillin/streptomycin.After 7 days, we performed a K  influx assay.

RNA purification, sequencing, and analysis
To perform poly-A RNA-sequencing, L1-L5 mouse DRG was homogenized with an RNeasy Mini Kit (Qiagen, Valencia, CA) using on-column DNase-I digestion according to the manufacturer's protocol.
RNA quality was assessed using a Nano-drop (Thermo Scientific) and RNA sequencing service was requested from LC science (Houston, TX).For the mouse DRG RNA-seq analysis, paired-end sequencing was performed with a read length of 150 bp.Read counts per gene per sample were quantified using HTSeq version 0.9.1.The detected total gene number is 25.091-25.664 in DRG samples.GPCR mRNA expression levels in human DRGs were obtained from the human DRG expression quantitative trait loci study (5).In short, DRG expression levels were determined from total RNA extracts measured using Affymetrix' Human Transcriptome Array 2.0 (Affymetrix, Santa Clara), scanned with Affymetrix' GeneChip 3000 G7 Instrument System, then probe intensities were normalized in a standard fashion.

Lipid pull-down assay
Isolated membrane proteins were pre-adsorbed to uncoated control agarose beads (Vector labs); the unbound fraction was collected and incubated for 24 h at 4°C with lipid-coated agarose beads (6).After extensive washing, bound proteins were eluted from the lipid-coated beads, suspended in Laemmli buffer containing 2% SDS (w/v) and 0.3M β-mercaptoethanol (Sigma), and heated for 5 min at 95°C to dissolve proteins before separation on 4-20% polyacrylamide/SDS gels.The lysate proteins were detected by western blot.

Immortal human SGCs cell line generation and transfection
We stumbled upon cells proliferating beyond passage 12 that were discovered by chance after primary culture from a DRG donor (NDRI, Philadelphia, PA).WT or mutant GPR37L1 were transfected to SGCs cell line using TransIT®-LT1 reagent (Mirus Madison, WI, USA, Cat Bo.MR2300).The control and mutant GPR37L1 or KCNJ3 expression levels were tested by Real-time PCR method.

Mouse SGC cultures and SGC-neuron co-cultures
DRGs were collected from both WT and KO mice (8 weeks male) and were incubated in 1 mg/ml Collagenase/Dispase (Roche Diagnostics, Madison, WI, USA) at 37℃ for 60 min, with agitation at 100 RPM and then followed by incubation in 0.05% trypsin/EDTA for 10 min.The digestion enzymes were prepared in Dulbecco's modified Eagle medium/F12 with GlutaMAX (ThermoFisher).After incubation with 0.1% trypsin inhibitor and centrifugation (300G), the cell pellet was gently triturated in a neurobasal medium containing 0.5 µM glutamine.Dissociated DRG cells were seeded on non-coated culture dishes for 4 h.SGCs and neurons were included for co-culture by hand-shaking of the culture flasks gently for 5 to 10 min and then resuspended by replacing them with a new neurobasal medium.The attached glial-like cells were cultured in the DMEM/F12 medium containing 10% fetal bovine serum and 1% streptomycin/penicillin to promote cell growth and inhibition of differentiation.After 2 to 6 days in culture, the glial-like cells were differentiated by application of the serum-free neural basal medium.The neuronrich fraction was seeded on a Poly L lysine (Sigma) coated plate with a Neurobasal medium containing 2% B27 (ThermoFisher) (7).

Computer simulations
The protein sequence of human GPR37L1 was downloaded from the UniProt database (ID: O60883) in fasta format.The predicted topology for GPR37L1 in UniProt was used for seven-transmembrane alignments and long-loop identification.Template selection and homology modeling were performed using the automated modeling server GPCR-ModSim (8).Human EDNRB (PDB: 6IGK) was selected as a template and active conformation of the model was generated.All other loops were also refined by the Prime module.Ligands were drawn in the Maestro suite in 2D format and were structurally preprocessed using LigPrep from the Schrodinger Suite (Schrödinger Release 2018-4: LigPrep, Schrödinger, LLC, New York, NY, 2018.).Protonation at a physiological pH (7±2) and energy minimization were performed using the OPLS3 force field.To elucidate the binding mode of all ligands in the binding site of the homology model of hGPR37L1, docking studies were performed with the help of Autodock4 software (9).Before the docking, the hGPR37L1 structure was prepared using the AutoDock Tools 4 software.The stability and intra-molecular conformational changes of the protein and molecular dynamics simulations (MDS) were performed on a 1000 ns time scale for the protein-ligand complex and 1000 trajectory structures were recorded.Using the GPU-accelerated DESMOND software (10), the top-scored docking poses were subjected to solvent-explicit, all-atom MDS.The OPLS-2005 force field was used for model generation of the protein-ligand complex, energy minimization, and MDS.The protein was inserted into the POPC lipid bilayer and the full system was immersed in a periodic orthorhombic water box TIP3P.The NPT ensemble class was used with the temperature set to 300 K and pressure set to 1.01325 bar.The trajectory clustering method of Desmond was used to cluster 1000 trajectory structures into ten clusters based on atomic root mean square deviation (RMSD).Then 100 ns MDS was performed on the cluster 1 structure to assess the stability of the docking complex.2D structure was plotted using protter software.

Intraganglionic injection of siRNA, PTX, MaR1, and AAV-virus
Mice were placed in a prone position under isoflurane anesthesia and a 1 cm posterior longitudinal skin incision was made in the lumbar portion of the spine.After the plate of the ipsilateral L5 vertebra was carefully removed to expose the L5 spinal nerve and the L4-L5 DRGs, microinjection of 1-2 µl of the solution was administrated to L4 and L5 DRG using a Hamilton syringe (80030) connected to a glass micropipette (tip diameter 10-20 μm) (11).After injection, the wound was covered with a gelatin sponge solution to prevent leakage and then closed with a suture.The following reagents were injected with gelatin sponge.siRNA: 3 g of Gpr37l1-targeting siRNA or scramble RNA was mixed with PEI/RNA polyplexes (1 µl LNP102).MaR1 (100 ng) or PTX (100 ng) with PBS as a vehicle was injected with gelatin spongey (4 µl).AAV9-Fabp7-mouse Gpr37L1 (1 x 10 12 ) were injected (2 µl) with gelatin spongey.

ELISA assay
Mouse IL-1β levels were measured using a mouse IL-1β ELISA Kit (R&D, Cat No. MLB00C, Minneapolis, MN), Human IL1β were detected using a human IL-1 β ELISA Kit (Proteintech, Rosemont, IL; Cat No. KE00021) according to the manufacturer's instructions.Briefly, 100 μl of culture medium was added to each well and a standard curve was included for each experiment.

Flow cytometry
Mouse DRG tissues were dissociated with 1 mg/ml Collagenase/Dispase (Roche) in a shaking incubator for 90 min.The dissociated tissues were incubated in 10% FBS-supplemented DMEM media for 1 h for neutralizing of the enzymes.The dissociated cells were washed out using a PBS + 10 mM EDTA solution.

Prediction of protein stability
To predict the stability of the protein in mutant GPR37L1, we used web-based protein stability software in known or newly discovered mutations.Change instability from a mutant ( M) protein to a wild-type ( W) form was defined as the difference in the corresponding unwinding free energies ΔΔG of the two proteins.
A negative ΔΔG value indicates that the mutant protein is more unstable.To calculate the protein stability, we used 9 different web-based prediction servers and presented the ΔΔG value (12,13)

Real-time quantitative PCR
Real-time quantitative PCR (RT-qPCR) assays were conducted in cDNA samples obtained from a reversetranscription reaction, using the BioRad CFX96 system (BioRad).Total RNA from the DRG and SGC cultures was extracted using the Direct-zol RNA Miniprep Kit (Zymo Research, Irvine, CA), and 0.5-1 μg RNAs were reverse transcribed using the iScript cDNA Synthesis Kit (Bio-Rad).Specific primers, including the GAPDH control, were designed using IDT SciTools Real-Time PCR software.We performed gene-specific mRNA analyses using the CTX96 Real-Time PCR System (Bio-Rad).
Quantitative PCR amplification reactions contained the same amount of reverse transcription product (50 ng), including 7.5 μl of 2× iQSYBR Green Mix (Bio-Rad) and 100-300 nM forward and reverse primers,

Preparation of plasma membrane (PM) fraction
Plasma membrane (PM) fraction was prepared from mouse or human DRG tissues and HEK cells using Mem-PER Plus Kit (Thermo Scientific, Cat No. 89842) and manufacture recommend protocol after biotinylation/streptavidin-magnetic bead (Thermo Scientific, Cat No.88817) pull down.Tissues or cells were harvested and washed out using a cold cell washing buffer.The tissues/cells were lysed by cell permeabilized buffer containing protease inhibitor (Sigma, Cat No. P8380) and phosphatase inhibitor (Sigma, Cat No.524631) using mechanical dissociation and ultrasonication.The lysate was centrifugated and the pellet was resuspended with PER lysis buffer containing protease and phosphatase inhibitor cocktail (Sigma).

Western blot
Protein samples were prepared from transfected cells, lipid pull-down beads, and DRGs.Tissue and cells were placed on ice and lysed with ice-cold RIPA buffer (Sigma, Cat No. R0278) with Protease Inhibitor Cocktail Tablet (pH 7.4) (Roche Diagnostics, Cat No. 5892988001).The cell lysates were centrifugated to remove insoluble debris and the protein level was detected via BCA assay.The supernatant was mixed with 4x Laemmli buffer (BioRad, Cat No. 1610747) and boiled for 10 min (100 ℃, containing BME for reducing Bis-Tris PAGE).Protein samples were electroporated on 4-20% gradient SurePAGE™ Bis-Tris gel (Genscript) or 4-20% Tris-glycine gel (BioRad) and blotted on a PVDF membrane (BioRad).Ponceau S staining was used for the detection of total proteins.The primary antibody was incubated with 1% BSA at 4℃ overnight.We used the following primary antibodies: anti-GPR37L1 antibody (Bioss, Rabbit,

Immunohistochemistry
Mice were deeply anesthetized with isoflurane and perfused through the ascending aorta with PBS, followed by 4% PFA.After the perfusion, L4-L5 DRGs were removed from the mice and post-fixed in the same fixative overnight.Human DRGs were obtained from the National Disease Research Interphase (NDRI) and fixed in 4% PFA overnight.The samples were then dehydrated with a 10% to 30% sucrose gradient, embedded in Tissue-Tek O.C.T., and sliced into sections (14 μm) in a cryostat.The sections were blocked with 5% donkey serum and 0.2% triton X-100 for 1 h at room temperature, then incubated overnight at 4 °C with the following primary antibodies: anti-GPR37L1 antibody (Alomone Labs, Rabbit, 1:500, Cat No. AGR-050), anti-GS antibody (Novus Biologicals, Rabbit, 1:1000, Cat No. NBP2-32241), anti-FABP7 (Neuromics, Mouse,1:1000, Cat No.MO22188), and anti-KCNJ10/Kir4.1 antibody (Thermo Scientific, Guinea pig, 1:800, Cat No. PA5-111798).The sections were washed in PBS and incubated with the following secondary antibodies (1:500, Biotium) for 2 h at room temperature: CF633-donkey antirabbit (Cat No. 20125), CF568 donkey anti-guinea pig (Cat No. 20377), and CF488A donkey anti-mouse (Cat No. 20952).For clarity, channel colors were exchanged in the presented mouse DRG images without disrupting the signal using ImageJ.DAPI (1:1,000, Thermo Scientific, Cat No. 62248) was used to stain the cell nuclei in tissue sections.The sections were then washed with PBS, mounted in Fluoromount G mounting medium (Southern Biotech, Cat No. 0100-01), and observed under a confocal laser scanning microscope (SP5 Inverted confocal-LSRC, Leica Microsystems).Some stained sections were also examined with a Leica SP5 or Zeiss 880 confocal microscope with Z-stack and Tile Scan.The maximum projected and stitched images were produced using the Zeiss Zen software or with ImageJ.

Patch-clamp recordings in mouse SGCs in whole-mount DRG preparation
Under urethane anesthesia, mice were rapidly euthanized, followed by careful isolation of lumbar DRGs placed in the oxygenated artificial cerebral spinal fluid (ACSF).DRGs were briefly digested (20 min) using an enzymatic mixture consisting of collagenase A (1 mg/mL) and Trypsin (0.25% original solution).
Intact DRGs were then incubated in ACSF oxygenated with 95% O2 and 5% CO2 at 34°C.Following incubation, DRGs were transferred to a recording chamber and continuously perfused (~3 ml/min) with ACSF.SGCs in whole mouse DRG could be visualized using a 40x water-immersion objective on an Olympus BX51WI microscope.The round or fusiform-shaped cell bodies of SGCs have small sizes (<10 μm) but are visible near the edges of DRG neurons.Patch pipettes (Chase Scientific Glass Inc.) were pulled and filled with a pipette solution containing (in mM): 126 potassium gluconate, 10 NaCl, 1 MgCl2, 10 EGTA, 2 Na-ATP, and 0.1 Mg-GTP, adjusted to pH 7.3 with KOH.The resistance of pipettes was 10-12 MΩ.A Whole-cell patch-clamp configuration was made on SGCs at room temperature using a Multiclamp 700B amplifier (Axon Instruments, Union City, CA).Under voltage clamp, at a holding potential of -80 mV, inward or outward currents were triggered by voltage steps beginning at -160 mV and increasing by 20 nM every 200 ms to a maximum of +40 mV (14).To isolate the inwardly rectifying potassium current (Kir), potassium currents from the same cells were recorded in the absence and presence of 100 μm extracellular barium for the blockade of the Kir4.1 channel (15,16).The Kir4.1 currents were obtained by digitally subtracting those currents in the absence and presence of barium.

Ti  influx assay
Ti  flux assay was conducted using the manufacture's and standard protocol (17).For examining Ti  flux in GPR37L1, KCNJ3, KCNJ3/KCNJ10-expressing Hek293 cells, and WT/E296K-GPR37L1-expressing human SGCs, cells were dislodged from tissue culture flasks using TrypLE Express reagent (Gibco, Cat No. 12604013) and transferred to a 50 mL centrifuge tube, and centrifuged at 500g for 2 min.The supernatant solution was removed by aspiration, and the pellet was resuspended at a concentration of 1000 cells/μL in a cell culture medium.Then 100 μl/well of the cell suspension was transferred to 96-well plates.
The seeded cells were incubated overnight in a humidified 5% CO2 cell culture incubator at 37 °C in low FBS media (2%).After overnight incubation, the cell culture medium was replaced with thallium influx dye-loading solution consisting of assay buffer (Hanks Balanced Salt Solution + 10 mM HEPES), 0.04% (w/v) Pluronic F-127 (Sigma), 1 μM of Thallos-AM (Ion bioscience) and 1 x TRS solution.Following a 60 min incubation at room temperature, the dye-loading solution was replaced with HBSS + 2 mM Ca 2+ and + 2 mM Mg 2+ 90 µl/well of assay buffer, and the plates were loaded into an Infiniti-M200 pro reader (Tecan, Männedorf, Switzerland).Data were acquired at (excitation 480, emission 520 nm) for 30 s intervals, followed by the addition of 10 μL/well of test compounds (30 min for KCNJ3 cell line or 1 h for KCNJ10 cell line).After baseline measurement for 5 min, 10 l/well Ti  stimulation was injected (at a rate of 100 l/sec), and fluorescence was measured for 10 min.The Tl  stimulus buffer consisted of 125 mM KHCO3, 1.8 mM CaSO4, 1 mM MgSO4, 5 mM glucose, 5 mM Tl2SO4, 10 mM HEPES, pH 7.4.The cover-glassed cultures of mouse and human SGCs were imaged using an Orca6 camera (Hamamatsu) every 3 sec with excitation at 480 nm and emission at 520 nm filters after thallium influx dye-loading solution.The imaging was performed using the MetaFluor software (Molecular Devices) in HBSS buffer supplemented with 2 mM Mg 2 and 2 mM Ca 2 .The activity was measured following stimulation with 500 µM Ti  solution.

Genome-wide association
Genome-wide association tests were performed in the large UK Biobank project cohort comprised of half a million participants (41, 67).Information about chronic pain reports collected at the visit was available at eight body sites: head (headaches), face, neck/shoulder, stomach/abdominal, back, hip, knee, and widespread.Control subjects were those that answered, "none of the above" at field 6159 for the question "In the last month have you experienced any of the following that interfered with your usual activities?".
Case subjects for chronic pain at body site X (among the eight listed) were those that answered "yes" to the question "Have you had pains for more than 3 months?";fields are 3571 (back), 4067 (face), 2956 (widespread), 3799 (headaches), 3414 (hip), 3773 (knee), 3404 (neck/shoulder), and 3741 (stomach/abdominal).Input genotyping data for association tests were from whole-exome sequencing (in VCF format), available for 200K individuals.Variants were annotated to GRCh38 reference provided by UK Biobank (http://biobank.ndph.ox.ac.uk/ukb/refer.cgi?id=3803) with Ensembl Variant Effect Predictor (VEP, release 99) (68).Rare coding variants were tested, and we required a minor allelic frequency of less than 0.01 and a minor allele count greater than 5.We used SAIGE (version 0.44.2) (69) to perform the tests, as it considers cryptic relatedness and guards against false-positive associations in the face of caseto-control imbalance (due to the cross-sectional nature of the cohort) by means of the saddle point approximation method (70).Covariables were: age, age squared, sex, genotyping platform, recruitment centers, and the first 40 principal genetic components.Subjects were of White British origin (field 22006).
Discarded subjects were those that opted out of the study, failed genotyping or imputation quality controls, or displayed sex aneuploidy or mismatched declared / genetically-determined sex.Retained SNPs displayed at least five counts of the minor allele in selected cases or control subjects.The meta-analysis at a given variant position across all eight chronic pain sites was performed using the inverse variancebased weighting scheme (71).A combined Annotation Dependent Depletion score (CADD v1.6) (43) was used to estimate the deleteriousness of the variants.It computed PHRED-scale CADD scores for variants, which provided a relative order compared to all existing variants in the human genome.Higher CADD scores for higher risk of deleteriousness.