GPR183-Oxysterol Axis in Spinal Cord Contributes to Neuropathic Pain

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

GPR183-oxysterol axis in spinal cord contributes to neuropathic pain

Kathryn Braden 1,2, Luigino Antonio Giancotti 1,2, Zhoumou Chen 1,2, Chelsea DeLeon 3, Nick

Latzo 3, Terri Boehm 1, Napoleon D’Cunha 1, Bonne M. Thompson 4, Timothy M. Doyle 1,2,

Jeffrey G. McDonald 4, John K. Walker 1,2, Grant R. Kolar 2,5, Christopher Kent Arnatt 2,3, Daniela

Salvemini 1,2

1 Department of Pharmacology and Physiology, Saint Louis University School of Medicine Downloaded from 2 Henry and Amelia Nasrallah Center for Neuroscience, Saint Louis University

3 Department of Chemistry, Saint Louis University

4 Center for Human Nutrition, University of Texas Southwestern Medical Center jpet.aspetjournals.org 5 Department of Pathology, Saint Louis University School of Medicine

at ASPET Journals on September 27, 2021

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Running Title: GPR183-oxysterol contribution to neuropathic pain

Corresponding author: Daniela Salvemini, Saint Louis University School of Medicine, 1402

South Grand Boulevard, Saint Louis, MO 63104. Phone: 1-314-977-6430, Fax: 1-314-977-6411,

Email: [email protected], ORCID ID: 0000-0002-0612-4448.

Text pages: 20

Tables: 0 Downloaded from

Figures: 6, 1 supplemental

References: 48 jpet.aspetjournals.org

Abstract word count: 232

Intro word count: 691 at ASPET Journals on September 27, 2021

Discussion word count: 1400

Non-standard abbreviations (A-Z): 25-OHC: 25-hydroxycholesterol, 7α,25-OHC: 7α,25- dihydroxycholesterol, CCI: Chronic constriction injury, CH25H: cholesterol 25-hydroxylase,

CNS: Central nervous system, CYP7B1: 25-hydroxycholesterol 7α-hydroxylase, DH-SC: dorsal- horn spinal cord, DMAP: 4-Dimethylaminopyridine, DRG: dorsal root ganglion, EAE: experimental autoimmune encephalitis, EBI2: Epstein-Barr Virus-Induced gene 2, ERK:

Extracellular signal-related kinase, HSD3B7: hydroxyl-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 7, i.th.: intrathecal, MS: multiple sclerosis, NMR: Nuclear magnetic resonance, PWT: Paw withdrawal threshold, RNA-Seq: RNA-Sequencing, SAR: Structure- activity relationship, SRE: Serum response element, TLC: Thin-layer chromatography

Recommended Section: Drug Discovery and Translational Medicine

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Abstract:

Neuropathic pain is a debilitating public health concern for which novel non-narcotic therapeutic targets are desperately needed. Using unbiased transcriptomic screening of the dorsal horn spinal cord after nerve injury we have identified that Gpr183 (Epstein-Barr induced gene 2

[EBI2]) is upregulated after chronic constriction injury (CCI) in rats. GPR183 is a chemotactic receptor known for its role in the maturation of B cells and the endogenous ligand is the

oxysterol, 7α,25-dihydroxycholesterol (7α,25-OHC). The role of GPR183 in the central nervous Downloaded from system is not well characterized and its role in pain is unknown. The profile of commercially available probes for GPR183 limits their use as pharmacological tools to dissect the roles of this receptor in pathophysiological settings. Using in silico modeling, we have screened a library of 5 jpet.aspetjournals.org million compounds to identify several novel small-molecule antagonists of GPR183 with nanomolar potency. These compounds are able to antagonize 7α,25-OHC-induced calcium at ASPET Journals on September 27, 2021 mobilization in vitro with IC50 values below 50nM. In vivo intrathecal injections of these antagonists during peak pain after CCI surgery reversed allodynia in male and female mice.

Acute intrathecal injection of the GPR183 ligand, 7α,25-OHC in naïve mice induced dose- dependent allodynia. Importantly, this effect was blocked using our novel GPR183 antagonists, suggesting spinal GPR183 activation as pro-nociceptive. These studies are the first to reveal a role for GPR183 in neuropathic pain and identify this receptor as a potential target for therapeutic intervention.

Significance Statement:

We have identified several novel GPR183 antagonists with nanomolar potency. Using these antagonists, we have demonstrated that GPR183 signaling in the spinal cord is pro-nociceptive.

These studies are the first to reveal a role for GPR183 in neuropathic pain and identify it as a potential target for therapeutic intervention.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Introduction:

Neuropathic pain conditions arising from nervous system injuries due to trauma, disease (i.e., diabetes) or neurotoxins (i.e. chemotherapy) are widespread and can be severe, debilitating and difficult to treat (Finnerup et al., 2015). Opioids are widely used to treat chronic pain but are limited by severe side effects and strong abuse liability (Toblin et al., 2011). Therefore, there is a high priority for developing novel non-opioid based analgesics. Downloaded from We used an unbiased transcriptomic analysis in this study to identify novel pathways and therapeutic targets in the development of neuropathic pain. In a model of traumatic nerve-injury induced neuropathic pain our findings identified GPR183/EBI2 (Epstein-Barr virus-induced gene jpet.aspetjournals.org 2) as a potential target for neuropathic pain. GPR183 was originally identified in B-cells as the most upregulated gene in response to Epstein-Barr virus infection (Birkenbach et al., 1993).

This receptor has been found in multiple human tissues including brain but was found most at ASPET Journals on September 27, 2021 abundantly in lymphoid organs and is most highly expressed on B-cells (Rosenkilde et al.,

2006). Similar patterns of expression have been found in rodents (Lein et al., 2007) and the human receptor sequence shares 88% homology with the rodent sequences, according to NCBI

Blast (Boratyn et al., 2013). GPR183 is important for the positioning of immune cells, particularly

B-cells, within lymphoid tissues, such as the spleen, for the launching of T-cell dependent antibody responses (Gatto et al., 2009; Pereira et al., 2009). GPR183 knockout mice are viable and have a normal gross phenotype: these mice have normal numbers of B cells and T cells with no defect in B-cell localization within the spleen (Pereira et al., 2009). Besides its role in regulating immune cell migration, GPR183 has been linked to metabolic diseases, multiple sclerosis, and cancer; accordingly, GPR183 has been proposed to represent a potential target for several diseases ranging from inflammation to cancer (Sun and Liu, 2015).

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

The primary endogenous ligand for GPR183 is the oxysterol, 7α,25-dihydroxycholesterol

(7α,25-OHC) (Hannedouche et al., 2011; Liu et al., 2011). This oxysterol is produced by the hydroxylation of cholesterol by cholesterol 25-hydroxylase (CH25H) resulting in its precursor,

25-hydroxycholesterol (25-OHC), which is subsequently hydroxylated by 25-hydroxycholesterol

7α-hydroxylase (CYP7B1) to form 7α,25-OHC (Mutemberezi et al., 2016). This oxysterol is rapidly degraded by hydroxyl-delta-5-steroid dehydrogenase, 3 beta- and steroid delta- isomerase 7 (HSD3B7) to be further metabolized into bile acids (Mutemberezi et al., 2016). Downloaded from 7α,25-OHC has proven to be a potent agonist of GPR183 in vivo (Liu et al., 2011; Chalmin et al., 2015; Lu et al., 2017; Wanke et al., 2017).

The role of GPR183 in the central nervous system (CNS) is still under investigation and its role jpet.aspetjournals.org in the context of pain is not known. At the cellular level, GPR183 has only been reported to be expressed in astrocytes within the CNS (Rutkowska et al., 2015). Other studies have found microglia are able to produce and release 7α,25-OHC, but it was not explored whether microglia at ASPET Journals on September 27, 2021 respond to the GPR183 ligand (Mutemberezi et al., 2018). GPR183 is Gαi-coupled and when activated by 7α,25-OHC, it can inhibit adenylate cyclase activity, increase phosphorylation of extracellular signal-regulated kinase (ERK) and p38 and trigger serum response element (SRE) activity (Rosenkilde et al., 2006; Benned-Jensen et al., 2011; Hannedouche et al., 2011; Liu et al., 2011; Benned-Jensen et al., 2013). Activation of these pathways in CNS glia as well as in dorsal root ganglia (DRG) neurons is crucial to the persistent pain sensitization and to pain chronification (Ji et al., 2009; Gomez et al., 2018).

Commercially available GPR183 antagonists are limited to a couple of compounds with chemical properties that question in vivo utility (Ardecky et al., 2010).To address this void, we embarked on drug discovery efforts to identify novel GPR183 antagonists for use in exploring the roles of GPR183 in neuropathic pain states. Our efforts led to the identification of several potent small-molecule selective GPR183 antagonists that were active in a rodent model of

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

neuropathic pain caused by chronic constriction of the sciatic nerve (Bennett and Xie, 1988).

These studies are the first step in investigating the role of GPR183 in neuropathic pain while providing the framework for future structure-activity relationship (SAR) investigations in the development of highly selective GPR183 antagonists. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Materials and Methods:

Experimental Animals:

Male and female ICR mice (8-9 weeks old; 25-40g starting weight) or Sprague Dawley rats (8-9 weeks old; 250g starting weight) from Envigo-Harlan Laboratories (Indianapolis, IN) were housed two to four per cage (rats) or five to ten per cage (mice) in a controlled environment (12 hour light/dark cycle) with food and water available ad libitum. All experiments were performed

with experimenters blinded to treatment conditions. All experiments were performed in Downloaded from accordance with the guidelines of the International Association for the Study of Pain and the

National Institutes of Health and approvals from the Saint Louis University Animal Care and Use

Committee. Experiments were performed in both male and female rodents; similar results were jpet.aspetjournals.org obtained in both sexes, so data was combined.

Test Compounds: at ASPET Journals on September 27, 2021 7α,25-dihydroxycholesterol (7α,25-OHC) was purchased from Avanti Polar Lipids (Alabaster,

AL) and dissolved in DMSO as a 2mM stock. For injections 7α,25-OHC was diluted in saline.

NIBR189 was purchased from Tocris Bioscience (Minneapolis, MN) and dissolved in DMSO as a 23mM stock. For intrathecal injections, NIBR189 was diluted in saline. Fluoronated 7α,25-

OHC analog (SLUPP-1492) was synthesized as described previously (Deng et al., 2016)

(detailed methods below and continued in Supplemental methods) and prepared as a 10mM stock in DMSO. For injections SLUPP-1492 was diluted in saline. SAE compounds were purchased from Enamine (Monmouth Jct., NJ), all compounds were purified via normal phase chromatography and had purities of ≥95%. 100mM Stock solutions of the SAE compounds were prepared in DMSO. For intrathecal injections, SAE compounds were diluted in saline. Acute intrathecal (i.th.) injections of compounds were performed as described previously (Lu and

Schmidtko, 2013), all compounds were administered intrathecally in a total volume of 5-10μL.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Chronic constriction injury (CCI) model:

CCI of the left sciatic nerve of mice and rats was performed under general anesthesia as previously described (Yosten et al., 2020). Briefly, animals were anesthetized with 2% isoflurane/O2, the left thigh was shaved and disinfected with Dermachlor solution. A small incision (1-1.5cm) was made in the middle of the lateral aspect of the left thigh to expose the sciatic nerve. The nerve was loosely ligated around the diameter at 3 distinct sites (1mm apart) using silk sutures (6.0, mice; 4.0, rats). The surgical site was closed with a skin clip and Downloaded from disinfected and treated with topical lidocaine (2%). Sham animals underwent the same procedure without nerve ligation. Peak allodynia developed by D7-D10 following CCI surgery.

Behavioral Testing: jpet.aspetjournals.org

Mechano-allodynia was measured as previously described (Yosten et al., 2020) using calibrated von Frey filaments (Stoelting; range in mice: 0.07-2.00g; in rats: 2-26g) using the Dixon up-and- down method (Dixon, 1991). Mechano-allodynia was defined as a significant (p<0.05) reduction at ASPET Journals on September 27, 2021 in mechanical paw withdrawal threshold [PWT (g)] compared to baseline forces (before treatment).

Cold allodynia was measured as previously described using the acetone test (Xing et al., 2007;

Yosten et al., 2020). Briefly, a small drop of acetone was applied to the hind paw of the animal using a flattened polyethylene tube and a syringe and the response to the cold stimulus was scored (0, no response; 1, brisk withdrawal or flick of the paw; 2, repeated flicking of the paw; 3, repeated flicking and licking of the paw). The test was repeated 3 times with an interval of 5 minutes between each application for each paw and the scores for each paw were summed and reported as the response score (maximum of 9). When mechanical and cold allodynia were measured on the same animal, mechano-allodynia was measured first with at least 15 minutes before testing cold allodynia.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Estrus smears:

Vaginal smears were taken for female mice and rats 5-7 days before experiment and after experiment until animals were sacrificed to confirm animals were cycling, and treatment did not alter their cycle. Cells were placed on a glass slide and allowed to dry, stained with Accustain

(MilliporeSigma; Saint Louis, MO) for 45 seconds and rinsed, as previously described (Byers et al., 2012). Fixed cells were viewed under a light microscope to determine their stage of estrus cycle. All animals displayed normal estrus cycles. Downloaded from

RNA-Sequencing:

On D10 after CCI, animals were perfused (1X PBS) under deep anesthesia and lower lumbar spinal cord was harvested and placed in RNA-Later. Total RNA was isolated using RNeasy Plus jpet.aspetjournals.org

Universal Mini kit (Qaigen; Germantown, MD) according to manufacturer’s protocols. RNA-

Sequencing was performed in the Saint Louis University Genomics Core Facility, Total RNA samples were quality assessed using an Agilent Bioanalyzer RNA Nano chip and were at ASPET Journals on September 27, 2021 determined to have an RNA integrity number of ≥9. Ribosomal RNA was depleted from total

RNA using the Eukaryotic RiboMinus Core Module v2 (Life Technologies, Thermofisher) and libraries were constructed using the Ion Total RNA-seq v2 kit (Life Technologies, Thermofisher) according to the manufacturer’s protocols. Sequencing was performed on an Ion Torrent Proton with a mean read length of ~140 nucleotides. Reads were aligned to the rat genome sequence

(version rn6) using the TMAP (Torrent Mapping Program) aligner (Homer, 2011) map4 algorithm, requiring a minimum seed length of 20 nucleotides and allowing soft-clipping at both

5’ and 3’ ends to accommodate spliced reads. The nucleotide coverage for all non-redundant exons was calculated and normalized to total exon coverage using BEDTools (Quinlan and Hall,

2010) and custom scripts in R (Team, 2015). Expression values are given as total normalized nucleotide exon coverage per gene. Fold changes in gene expression and p values were calculated using R and Microsoft Excel.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

RNAScope®:

On D10 after CCI surgery animals were perfused under deep anesthesia and with 4% paraformaldehyde and lower lumbar spinal cord was harvested and post-fixed in 4% paraformaldehyde. Spinal cord lumbar sections were cryosectioned at 10 micrometers and stained using the RNAscope® technique. Probes for rat Gfap (NM_017009.2, Probe- Rn-Gfap-

C2), Aif1 (NM_017196.3, Probe- Rn-Aif1-C3), Rbfox3 (NM_001134498.2, Probe- Rn-Rbfox3-

C2), and Gpr183 (NM_001109386.1, Probe- Rn-Gpr183) were incubated with tissue strictly Downloaded from according to the Manual RNAscope® Fluorescence Multiplex Protocol (v2) (Advanced Cell

Diagnostics, Newark, CA). Sections were imaged as Z-stacks of the central lamina 1 and 2 region of the dorsal horn on a Lecia TCS SP8 confocal microscope using a 40x (NA 1.30) (Leica jpet.aspetjournals.org

Microsystems, Buffalo Grove, IL). RNAscope signal was dilated by 3 pixel diameters for large overview display purposes or else signal was dilated by 0.5 pixel diameters. For analysis,

RNAscope signal was dilated for each slice by 0.5 pixel diameters and Gpr183 was isolated by at ASPET Journals on September 27, 2021 thresholding the channel on positive signal, setting a selection area, and applying it to the appropriate lineage marker (GFAP, Aif1, or Rbfox3) within FIJI (PMID 22743772) using a custom macro (Supplementary methods). Particle counts of the signals were made in the resulting overlapping region (which represented the major RNA pool of a particular cell) for each slice.

Oxysterol Quantification:

Male Sprague Dawley rats underwent CCI surgery on D0 and were taken down on either D0,

D5, or D11 after surgery. Rats were perfused under deep anesthesia with 1X PBS, ipsilateral dorsal horn spinal cord was harvested and flash-frozen in liquid nitrogen. Oxysterols were measured according to McDonald et. al. (McDonald et al., 2012).

In silico modeling:

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

An inactive homology model of GPR183 based largely upon the Chemokine Receptor CCR5

(5UIW) was downloaded from the GPCRdb. (Pándy-Szekeres et al., 2017) Using Schrödinger, protein preparation was run to minimize the energy of the protein (OPLS3 force field).

Schrödinger Phase was then employed to build a pharmacophore hypothesis built around the binding of NIBR189 and GSK682753A in GPR183 using the automated build function within

Phase for receptor-ligand complex pharmacophore. This included features such as: aromatic, hydrophobic, H-bond acceptor/donator, and size exclusion spheres. Using a freely available Downloaded from database from Enamine (Enamine_Diverse_REAL_drug-like_5M library), 5 million compounds were screened for their “likeness” to the properties of the NIBR189 and GSK682753A separately. The top ten thousand screening hits from each screen based upon their feature jpet.aspetjournals.org matching with the two pharmacophores were chosen for further screening. The twenty thousand screening hits were then used in a high-throughput GLIDE docking and the top 800 compounds

(4%) with the lowest GlideScores (computational estimate of binding) were chosen for more at ASPET Journals on September 27, 2021 precise docking studies. The top 800 compounds (4%) with the lowest GlideScores were subjected to standard precision docking using flexible ligand sampling in GLIDE using the default settings. The compounds were then sorted based upon their predicted Log(S) values (all the compounds were prescreened to have a Log(P) of less than 5). Compounds with Log(S) values larger than -4 were then sorted based upon their GlideScores and the top 16 commercially available compounds in that pool were then ordered from Enamine for initial in vitro screening.

Synthesis of SLUPP-1492:

Material: The 25-hydroxycholesterol was purchased from ChemShuttle (Hayward, CA), all other reagents and solvents were purchased from Sigma-Aldrich (St. Louis, MO), Alfa Aesar (Ward

Hill, MA), or J.T. Baker (Radnor, PA) and used as received.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Instrumentation: The purities of the final compounds were characterized by high-performance liquid chromatography (HPLC) using a gradient elution program (Ascentis Express Peptide C18 column, acetonitrile/water 5/95 -> 95/5, 5 min, 0.05% trifluoroacetic acid) and UV detection (245 nM). TLC was performed on Merck KGaA TLC silica gel 60 F254 plates. Visualization was accomplished by using phosphomolybdic acid solution followed by heat and by UV fluorescence

1 (λmax 254 nm). H nuclear magnetic resonance (NMR) was obtained on a Bruker 400 MHz instrument and all chemical shifts are referenced to residual solvent peaks (details in Downloaded from supplemental methods).

Cell line and culture: jpet.aspetjournals.org

The Human leukemia (HL)-60 cells (American Type Culture Collection; Manassas, VA) were cultivated in RPMI 1640 media containing 10% heat-inactivated fetal bovine serum (FBS), 1% penicillin, and 1% GlutaMax. Cells were passaged every 3 days and maintained at a cell at ASPET Journals on September 27, 2021 concentration below 1x106 to prevent differentiation. The cells were incubated at 37C under 5%

CO2.

Calcium mobilization assays:

Serial dilutions (5X) of compounds were prepared in 50:1 HBSS/HEPES.HL-60 cells (1x107) were incubated in 50:1 HBSS/HEPES containing 5 M indo-1-AM (Thermo Fischer Scientific;

Waltham, MA) and 0.05% pluronic acid for 0.5 hours at room temperature. Cells were centrifuged and washed with 50:1 HBSS/HEPES and resuspended in buffer. Cells were loaded onto a black 96-well Greiner Bio-One (Thermo Fischer Scientific; Waltham, MA) clear-bottom plate at 100,000 cells/well for a 5X dilution. For agonism assays, cells were immediately incubated for 15 minutes at 37C inside a FlexStation 3 Multimode Plate Reader (Molecular

Devices; Sunnyvale, CA). Following the 15 minute incubation period, 5X of the compound was added and fluorescence was read for 150 seconds at 37C. This method was utilized to

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

determine the EC80 value of 7α,25-OHC (EC8o = 200 nM). For antagonism assays, before the addition of agonist, 5X of the antagonist was added and incubated for 15 minutes at 37C inside a FlexStation3. After allowing 15 minutes of equilibration, the determined EC80 of 7,25-OHC

(200 nM) was added, and fluorescence was read for 150 secs at 37C. Calcium mobilization was determined ratiometrically using ex 350 nm and em 405/490 nm. Dose-response data were normalized to a 0.5% DMSO vehicle control and the maximum response. From the normalized data, non-linear regression curves were then generated to calculate the appropriate EC50 and Downloaded from

IC50 values. Each compound was run in triplicate (n=3). siRNA knock-down of GPR183: jpet.aspetjournals.org siRNA was purchased from Santa Cruz Biotechnology (Dallas, TX). HL-60 cells were centrifuged and counted with a hemocytometer to obtain 2.5x106 cells. Transfection of the siRNA was performed with Lipofectamine 2000 reagent (Invitrogen; Calsbad, CA) and Opti- at ASPET Journals on September 27, 2021 MEM. Lipofectamine and siRNA (200pmol) were incubated at room temperature for 15 minutes before addition to 5 mL of cell suspension in media. Transfected cells were incubated at 37C under 5% CO2 for 48 hours. HL-60 control cells were treated with Lipofectamine 2000 reagent without siRNA. After 48 hours cells were utilized in calcium mobilization assays.

Statistical analysis:

Data is expressed as mean ± SD or SEM for n biological replicates and analyzed by paired or unpaired t test or one-way or two-way repeated measures ANOVA with Dunnett’s multiple comparisons. Sphericity was tested with Mauchly’s test and Greenhouse-Geiser corrections were used when necessary. Significant differences were defined as p<0.05. Statistical analyses were performed using Graphpad Prism (versions 5.00-8.1.1. for Windows, GraphPad Software,

San Diego, CA, www.graphpad.com)

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Results:

GPR183 upregulation in spinal cord after nerve injury

A well-characterized rodent model of neuropathic pain was used. In this model, constriction of the sciatic nerve produces robust mechano-allodyina that peaks within 7-10 days after injury and last for several weeks (Bennett and Xie, 1988). RNA-Seq analysis of dorsal horn spinal cord tissues (DH-SC) harvested at peak mechano-allodynia (D10 after CCI surgery; Fig. 1A) revealed a 2.8-fold (p=7.59x10-14, FDR=9.98x10-12) increase in Gpr183 expression in the DH-SC Downloaded from from rats with CCI compared to those that received sham surgery (Fig. 1B). To confirm this upregulation and identify where Gpr183 is expressed, we used RNAscope®-based in situ hybridization performed in the spinal cord of rats with CCI. Gpr183 is expressed in the dorsal jpet.aspetjournals.org and ventral horn of the spinal cord (Supplementary Fig. S1). However, Gpr183 expression increased 2.4-fold (p=0.024; paired t-test) within lamina 1 and 2 of the DH-SC from mice with

CCI ipsilateral to the nerve injury (Fig. 1C-D). Further analyses revealed that while Gpr183 co- at ASPET Journals on September 27, 2021 localized in microglia, astrocytes and neurons (Fig. 2), its expression increased in microglia

(6.4-fold; p=0.0021) and astrocytes (2.5-fold; p=0.0021), but not neurons (p=0.4280) (Fig. 2) after CCI surgery.

Oxysterol metabolism in spinal cord after nerve injury

7α,25-OHC is the most potent ligand of GPR183, so if GPR183 were functionally involved in pain, we would expect a correlating increase in the ligand. So we performed mass spectrometry to quantify the oxysterol content of the spinal cord at different time points after CCI surgery in rats (day 0 pre-surgery, day 5 and day 11 post surgery). However, 7α,25-OHC was undetectable at all time points after CCI surgery (n=8). We were able to detect its precursor, 25- hydroxycholesterol (25-OHC) but levels did not change over time (from 0.16 ng/mg, to 0.12 ng/mg and 0.09 ng/mg on day 0, 5 and 11 respectively; one-way ANOVA p=0.0662 for n=8).

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Failure of NIBR189 to inhibit CCI-induced mechano-allodynia

The commercially available small molecule GPR183 antagonists are limited to NIBR189 (IC50

11nM) (Gessier et al., 2014) and GSK682753A (IC50=0.2μM) (Benned-Jensen et al., 2013).

Since GSK682753A has poor microsomal and plasma stability (Ardecky et al., 2010) and is therefore not suitable for in vivo studies, we employed NIBR189 as our first attempt to explore the role of GPR183 in nerve-injury induced neuropathic pain. Intrathecal administration of

NIBR189 on day 7 after CCI surgery in male mice failed to reverse mechano-allodynia at doses Downloaded from as high as 23μM (Fig. 3). These doses were chosen based on previous literature using

NIBR189 in vitro (Gessier et al., 2014; Preuss et al., 2014; Rutkowska et al., 2015). Since information regarding validation of NIBR189 as a specific GPR183 antagonist is limited to the jpet.aspetjournals.org original pharmacological characterization (Gessier et al., 2014), we undertook our own validation. We discovered that NIBR189 was unable to reliably inhibit 7α,25-OHC-induced calcium mobilization; we were unable to obtain an IC50 value for this compound. Due to at ASPET Journals on September 27, 2021 significant variations in dose-responses, we observed single plate IC50 values that consistently had standard deviations of multiple magnitudes larger/smaller. This viability was not resolved by different lots or manufacturers of NIBR189, different lots of cells, using different compound vehicles, changes in assay protocols, etc. Importantly, this variability was only observed for

NIBR189, and none of the other agonists or antagonists used in this study. NIBR189 was previously claimed to be a competitive GPR183 antagonist reported to inhibit 7α,25-OHC binding to GPR183 with an IC50 of 11nM (Gessier et al., 2014). Our findings do not support such a claim. In fact, we have found that it does not consistently inhibit 7α,25-OHC induced signaling, and overall, has an unclear pharmacology that is highly variable. The reason for this discrepancy is at present unclear. Therefore, we underwent drug discovery efforts to identify more GPR183 antagonists that could be used for in vivo proof of concept studies.

Discovery of potent and selective GPR183 antagonists with pharmacological activity

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Knowing that both NIBR189 and GSK682753A have shown binding affinity to GPR183, they were docked within a GPR183 homology model in order to build a pharmacophore model. The pharmacophore model based upon the NIBR189-GPR183 and GSK682753A-GPR183 bound structures was built using the automated pharmacophore builder in Schrodinger Phase. This model consisted of aromatic, hydrophobic, H-bond acceptor/donator, and size exclusion spheres in 3-dimensional space describing the key features needed for binding to GPR183 by those two compounds. Using an in silico approach, a library of 5 million compounds was Downloaded from screened for similarity to the GPR183 pharmacophore model. Compounds with the highest similarity scores were docked and ranked based upon their thermodynamics of binding (Fig.

4A). The top 16 commercially available compounds were purchased and then tested for jpet.aspetjournals.org

GPR183-specific agonism and antagonism in a calcium mobilization assay (Figs. 4B-D). Three of those compounds were able to antagonize 7α,25-OHC-induced calcium mobilization with IC50 values below 50 nM (Fig. 4B, D). These compounds were unable to effect calcium mobilization at ASPET Journals on September 27, 2021 in the HL-60 cells on their own. Results were confirmed to be GPR183-specific using siRNA to block protein expression. In these studies, 7α,25-OHC-induced calcium mobilization (Fig. 4E), and GPR183 antagonism (Fig. 4F) were both abolished using GPR183-specific siRNA.

GPR183 antagonists reverse CCI-induced mechanical allodynia

Using these lead compounds, we were able to assess whether GPR183 is functionally involved in neuropathic pain states. When administered in vivo to mice these GPR183 antagonists (SAE-

1, SAE-10, and SAE-14) were able to reverse CCI-induced mechanical allodynia in a time- dependent manner (Fig. 5). For our in vivo studies we increased the concentration to 100x the in vitro IC50 of these compounds as this is a reasonable approach when transitioning from in vitro to in vivo investigations. Future studies will include full dose responses with the most promising lead molecules that will emerge from our SAR studies.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

GPR183 signaling in naïve animals is pro-nociceptive

Our results using GPR183 antagonists in neuropathic pain models suggest that GPR183 signaling in the spinal cord contributes to the maintenance of neuropathic pain. We reasoned that if this was true then i.th. injections of the most potent GPR183 ligand, 7α,25-OHC should recapitulate behavioral consequences of neuropathic pain states (i.e., allodynia). Indeed, i.th. injections of 7α,25-OHC in mice induced a dose and time-dependent mechano-allodynia

(ED50=74nM; Fig. 6A) and cold allodynia (Fig. 6B). Additionally, the use of a fluorinated analog Downloaded from of 7α,25-OHC, SLUPP-1492, induced similar results (Figs. 6A-B). Importantly, pre-treatment with one of our novel GPR183 antagonists (SAE-14) was able to block these effects of 7α,25-

OHC in a dose-dependent manner (Figs. 6C-D). jpet.aspetjournals.org

at ASPET Journals on September 27, 2021

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Discussion:

Our results suggest that spinal activation of GPR183 by the oxysterol 7α,25-OHC is pronociceptive and that blocking GPR183 in the spinal cord can attenuate mechano-allodynia following traumatic nerve injury. Furthermore, we have shown that GPR183 co-localizes with microglia and astrocytes within the dorsal horn spinal cord, suggesting that the receptor is specifically upregulated on these cell types. This is consistent with GPR183’s well-characterized

immune function in the peripheral immune system. GPR183 has been previously shown to be Downloaded from expressed on astrocytes (Rutkowska et al., 2015) and have some function in neuroinflammatory states (Kurschus and Wanke, 2018). However, the significance of an upregulation of this receptor on astrocytes and microglia in the context of neuropathic pain is not yet known but will jpet.aspetjournals.org be examined in future studies as we develop lead GPR183 antagonists.

The previously developed GPR183 antagonist, NIBR189, proved ineffective at reversing at ASPET Journals on September 27, 2021 neuropathic pain states, and we were unable to independently validate its specificity toward

GPR183. Previous studies have consistently shown that it is able to prevent in vitro migration of

GPR183-expressing cells towards either exogenous 7α,25-OHC or media containing released

GPR183 ligands (Preuss et al., 2014; Rutkowska et al., 2015; Rutkowska et al., 2016b; Clottu et al., 2017; Wanke et al., 2017). However, only a few studies have investigated the functional effects of NIBR189 on the downstream signaling of GPR183 or in disease states. In human- derived macrophages, it was found that 7α,25-OHC-induced intracellular calcium mobilization which could be inhibited by NIBR189, but an IC50 was not calculated (Preuss et al., 2014). A few studies found NIBR189 is able to reduce 7α,25-OHC-induced changes in cell impedance of macrophages (Preuss et al., 2014), and astrocytes (Rutkowska et al., 2018). But this measurement is non-specific and only implies calcium signaling or migration and does not hone in on the molecular consequences of 7α,25-OHC, or NIBR189 treatment. In neonatal mouse cerebellar slices, it was shown that culture with either 7α,25-OHC or NIBR189 was able to

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

reduce myelination by 20 days, the authors of this study postulate that both the agonist and antagonist have similar effects due to 7α,25-OHC’s ability to cause receptor internalization, thus having a functionally antagonistic effect (Rutkowska et al., 2017). However, the authors of that study did not do confirmatory experiments of this hypothesis other than to show internalization of the receptor by 7α,25-OHC, and not NIBR189 (Rutkowska et al., 2017). It was previously shown that 7α,25-OHC is able to upregulate p-ERK, increase calcium flux, and induce migration of human astrocytes in vitro (Rutkowska et al., 2015). However only the upregulation of p-ERK Downloaded from was ameliorated by NIBR189 treatment while the calcium flux and migration responses were reduced in GPR183 knockout cells (Rutkowska et al., 2015). Many of these in vitro studies were conducted by the same group that characterized NIBR189, and thus did not independently jpet.aspetjournals.org validate the specificity of this compound. In vivo studies using NIBR189 are limited despite the original characterization claiming favorable pharmacokinetic properties (Gessier et al., 2014).

Acute systemic treatment with NIBR189 in naïve mice resulted in a disruption of dendritic cell at ASPET Journals on September 27, 2021 positioning within the T cell zone of the spleen (Lu et al., 2017). In GPR183 knockout mice the dendritic cells were still uniformly distributed within the T cell zone, however, the authors did not comment on this discrepancy as the GPR183 knockout mice had a significantly reduced number of dendritic cells and not all tissues had enough cells for comparative analysis (Lu et al., 2017).

In a disease model of orthotopic lung transplant, long-term NIBR189 treatment inhibited markers of chronic rejection, with similar results found in GPR183 knockout mice (Smirnova et al., 2019).

This study is the only one to use NIBR189 in a disease model and to find the same results in the knockout model. Based on this and our results using this compound, we believe that NIBR189 is not suitable for in vivo use. Therefore, we propose that the SAE compounds described in this study could be used to further characterize the in vivo properties of GPR183.

In our studies, we were unable to detect 7α,25-OHC in the spinal cord after nerve injury, but this was not surprising as this oxysterol is present in very small levels in biological fluids making its

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

detection very difficult despite sophisticated bioanalytical methodologies (Griffiths et al., 2016).

Nevertheless, it is well accepted that 7α,25-OHC is a very potent signaling molecule able to exert profound effects at very low levels (Liu et al., 2011; Chalmin et al., 2015; Lu et al., 2017;

Wanke et al., 2017).

The molecular mechanisms underpinning the analgesic effects provided by GPR183 antagonists will be explored in future studies as the chemistry of these antagonists evolves.

However, a few possibilities are offered based upon our understanding of this system in other Downloaded from areas. GPR183 has mostly been characterized as a chemotactic receptor in the peripheral immune system (Sun and Liu, 2015), so the downstream effects of its signaling have not been jpet.aspetjournals.org elucidated. GPR183 is known to be coupled to Gαi and when activated can induce phosphorylation of ERK, p38, or SRE (Rutkowska et al., 2016a). The activation of these signaling pathways in astrocytes or microglia within the spinal cord has been shown to be important for the initiation of neuroinflammation and the maintenance of chronic pain (Zhuang et at ASPET Journals on September 27, 2021 al., 2005). Therefore, it is possible that 7α,25-OHC-mediated GPR183 activation on CNS glia could be contributing to neuropathic pain states through neuroinflammatory mechanisms.

GPR183 has been previously investigated for a role in neuroinflammatory diseases such as multiple sclerosis (MS). However, in mouse models of MS two groups found conflicting evidence, with one showing that knockout of the rate-limiting enzyme for 7α,25-OHC production

(CH25H-/-) resulted in an exacerbated course of experimental autoimmune encephalitis (EAE)

(Reboldi et al., 2014) and another found the same CH25H-/- mice to have a delayed course of

EAE (Chalmin et al., 2015). In GPR183 knockout mice, there was no difference in EAE disease course compared to wild type mice (Wanke et al., 2017). However, when a transfer model of

EAE was used and GPR183-deficient Th17 cells were adoptively transferred to Rag-/- mice, they had a delayed onset of disease (Wanke et al., 2017), similar to what was seen in Chalmin and colleagues’ CH25H-/- mice (Chalmin et al., 2015). Therefore, it is likely that GPR183-oxysterol

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

signaling contributes to neuroinflammatory states such as multiple sclerosis through recruitment of infiltrating immune cells to the CNS. Infiltration of immune cells has also been shown to be a contributing factor to neuropathic pain states which can lead to enhancement of neuroinflammatory signaling and central sensitization (Grace et al., 2011). Future studies will reveal how oxysterols and GPR183 contribute to the recruitment of peripheral immune cells to the CNS following traumatic nerve injury. Importantly, others have identified GPR183 as being upregulated in the spinal cord of rats after a post-surgical model of pain (Raithel et al., 2018), Downloaded from supporting our data and further implicating GPR183 in pain. These findings further implicate

GPR183 in disease states involving neuroinflammation, including neuropathic pain. This provides a platform to further investigate the chemistry of GPR183 within the CNS and lead to jpet.aspetjournals.org more viable tools for the pharmacological characterization of this receptor.

Acknowledgments: We would like to thank Dr. Dale Dorsett and the Saint Louis University

Genomics Core for their assistance with the RNA Sequencing. We would also like to thank the at ASPET Journals on September 27, 2021

NMR facility and the Department of Chemistry at Saint Louis University for use of the 400 MHz

NMR.

Authorship Contributions:

Participated in Research Design: Braden, Arnatt, Salvemini

Conducted Experiments: Braden, Giancotti, Chen, DeLeon, Latzo, Boehm, D’Cunha,

Thompson, Doyle, Kolar

Contributed new reagents/analytical tools: McDonald, Walker, Kolar

Performed data analysis: Braden, Giancotti, DeLeon, Thompson, Doyle, Kolar, Arnatt

Wrote or contributed to writing of the manuscript: Braden, Doyle, McDonald, Walker, Kolar,

Arnatt, Salvemini

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Footnotes*† Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021

* This work has been presented in part at the following meetings:

Braden K, Giancotti L, Chen Z, DeLeon C, Latzo N, Doyle T, Kolar G, Walker J, Arnatt C, Salvemini D.

Identifying a role for GPR183 and 7α,25-dihydroxycholesterol in neuropathic pain. Poster, Henry and

Amelia Nasrallah Center for Neuroscience Research Symposium. Saint Louis, MO, November 1, 2019.

Braden K, Giancotti L, Chen Z, Salvemini D. GPR183 as a novel target for the treatment of neuropathic pain. Poster, Society for Neuroscience Meeting. Chicago, IL, October 19-23, 2019.

† All reprint requests should be addressed to: Daniela Salvemini, Saint Louis University School of

Medicine, 1402 South Grand Boulevard, Saint Louis, MO 63104. Email: [email protected]

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

References:

Ardecky R, Sergienko E, Zou J, Ganji S, Brown B, Sun Q, Ma CT, Hood B, Nguyen K, Vasile S,

Suyama E, Mangravita-Novo A, Salaniwal S, Kung P, Smith LH, Chung TDY, Jackson

MR, Pinkerton AB and Rickert R (2010) Functional Antagonists of EBI-2, in Probe

Reports from the NIH Molecular Libraries Program, Bethesda (MD).

Benned-Jensen T, Madsen CM, Arfelt KN, Smethurts C, Blanchard A, Jepras R and Rosenkilde

MM (2013) Small molecule antagonism of oxysterol-induced Epstein-Barr virus induced Downloaded from

gene 2 (EBI2) activation. FEBS Open Bio 3:156-160.

Benned-Jensen T, Smethurst C, Holst PJ, Page KR, Sauls H, Sivertsen B, Schwartz TW,

Blanchard A, Jepras R and Rosenkilde MM (2011) Ligand modulation of the Epstein- jpet.aspetjournals.org

Barr virus-induced seven-transmembrane receptor EBI2: identification of a potent and

efficacious inverse agonist. J Biol Chem 286:29292-29302.

Bennett GJ and Xie YK (1988) A peripheral mononeuropathy in rat that produces disorders of at ASPET Journals on September 27, 2021

pain sensation like those seen in man. Pain 33:87-107.

Birkenbach M, Josefsen K, Yalamanchili R, Lenoir G and Kieff E (1993) Epstein-Barr virus-

induced genes: first lymphocyte-specific G protein-coupled peptide receptors. J Virol

67:2209-2220.

Boratyn GM, Camacho C, Cooper PS, Coulouris G, Fong A, Ma N, Madden TL, Matten WT,

McGinnis SD, Merezhuk Y, Raytselis Y, Sayers EW, Tao T, Ye J and Zaretskaya I

(2013) BLAST: a more efficient report with usability improvements. Nucleic Acids Res

41:W29-W33.

Byers SL, Wiles MV, Dunn SL and Taft RA (2012) Mouse estrous cycle identification tool and

images. PLoS One 7:e35538.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Chalmin F, Rochemont V, Lippens C, Clottu A, Sailer AW, Merkler D, Hugues S and Pot C

(2015) Oxysterols regulate encephalitogenic CD4(+) T cell trafficking during central

nervous system autoimmunity. J Autoimmun 56:45-55.

Clottu AS, Mathias A, Sailer AW, Schluep M, Seebach JD, Du Pasquier R and Pot C (2017)

EBI2 Expression and Function: Robust in Memory Lymphocytes and Increased by

Natalizumab in Multiple Sclerosis. Cell Rep 18:213-224.

Deng X, Sun S, Wu J, Kuei C, Joseph V, Liu C and Mani NS (2016) Fluoro analogs of bioactive Downloaded from oxy-sterols: Synthesis of an EBI2 agonist with enhanced metabolic stability. Bioorg Med

Chem Lett 26:4888-4891.

Dixon WJ (1991) Staircase bioassay: the up-and-down method. Neurosci Biobehav Rev 15:47- jpet.aspetjournals.org

50.

Finnerup NB, Attal N, Haroutounian S, McNicol E, Baron R, Dworkin RH, Gilron I, Haanpaa M,

Hansson P, Jensen TS, Kamerman PR, Lund K, Moore A, Raja SN, Rice AS, at ASPET Journals on September 27, 2021

Rowbotham M, Sena E, Siddall P, Smith BH and Wallace M (2015) Pharmacotherapy for

neuropathic pain in adults: a systematic review and meta-analysis. Lancet Neurology

14:162-173.

Gatto D, Paus D, Basten A, Mackay CR and Brink R (2009) Guidance of B cells by the orphan

G protein-coupled receptor EBI2 shapes humoral immune responses. Immunity 31:259-

269.

Gessier F, Preuss I, Yin H, Rosenkilde MM, Laurent S, Endres R, Chen YA, Marsilje TH,

Seuwen K, Nguyen DG and Sailer AW (2014) Identification and characterization of small

molecule modulators of the Epstein-Barr virus-induced gene 2 (EBI2) receptor. J Med

Chem 57:3358-3368.

Gomez R, Kohler DM, Brackley AD, Henry MA and Jeske NA (2018) Serum response factor

mediates nociceptor inflammatory pain plasticity. Pain Reports 3:e658-e658.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Grace PM, Rolan PE and Hutchinson MR (2011) Peripheral immune contributions to the

maintenance of central glial activation underlying neuropathic pain. Brain Behav Immun

25:1322-1332.

Griffiths WJ, Abdel-Khalik J, Crick PJ, Yutuc E and Wang Y (2016) New methods for analysis of

oxysterols and related compounds by LC-MS. J Steroid Biochem Mol Biol 162:4-26.

Hannedouche S, Zhang J, Yi T, Shen W, Nguyen D, Pereira JP, Guerini D, Baumgarten BU,

Roggo S, Wen B, Knochenmuss R, Noel S, Gessier F, Kelly LM, Vanek M, Laurent S, Downloaded from Preuss I, Miault C, Christen I, Karuna R, Li W, Koo DI, Suply T, Schmedt C, Peters EC,

Falchetto R, Katopodis A, Spanka C, Roy MO, Detheux M, Chen YA, Schultz PG, Cho

CY, Seuwen K, Cyster JG and Sailer AW (2011) Oxysterols direct immune cell migration jpet.aspetjournals.org

via EBI2. Nature 475:524-527.

Homer N (2011) TMAP: The Torrent Mapping Program, in.

Ji RR, Gereau RWt, Malcangio M and Strichartz GR (2009) MAP kinase and pain. Brain Res at ASPET Journals on September 27, 2021

Rev 60:135-148.

Kurschus FC and Wanke F (2018) EBI2 - Sensor for dihydroxycholesterol gradients in

neuroinflammation. Biochimie 153: 52-55.

Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, Boe AF, Boguski MS,

Brockway KS, Byrnes EJ, Chen L, Chen L, Chen TM, Chin MC, Chong J, Crook BE,

Czaplinska A, Dang CN, Datta S, Dee NR, Desaki AL, Desta T, Diep E, Dolbeare TA,

Donelan MJ, Dong HW, Dougherty JG, Duncan BJ, Ebbert AJ, Eichele G, Estin LK,

Faber C, Facer BA, Fields R, Fischer SR, Fliss TP, Frensley C, Gates SN, Glattfelder

KJ, Halverson KR, Hart MR, Hohmann JG, Howell MP, Jeung DP, Johnson RA, Karr PT,

Kawal R, Kidney JM, Knapik RH, Kuan CL, Lake JH, Laramee AR, Larsen KD, Lau C,

Lemon TA, Liang AJ, Liu Y, Luong LT, Michaels J, Morgan JJ, Morgan RJ, Mortrud MT,

Mosqueda NF, Ng LL, Ng R, Orta GJ, Overly CC, Pak TH, Parry SE, Pathak SD,

Pearson OC, Puchalski RB, Riley ZL, Rockett HR, Rowland SA, Royall JJ, Ruiz MJ,

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Sarno NR, Schaffnit K, Shapovalova NV, Sivisay T, Slaughterbeck CR, Smith SC, Smith

KA, Smith BI, Sodt AJ, Stewart NN, Stumpf KR, Sunkin SM, Sutram M, Tam A, Teemer

CD, Thaller C, Thompson CL, Varnam LR, Visel A, Whitlock RM, Wohnoutka PE,

Wolkey CK, Wong VY, et al. (2007) Genome-wide atlas of gene expression in the adult

mouse brain. Nature 445:168-176.

Liu C, Yang XV, Wu J, Kuei C, Mani NS, Zhang L, Yu J, Sutton SW, Qin N, Banie H, Karlsson L,

Sun S and Lovenberg TW (2011) Oxysterols direct B-cell migration through EBI2. Nature Downloaded from 475:519-523.

Lu E, Dang EV, McDonald JG and Cyster JG (2017) Distinct oxysterol requirements for

positioning naive and activated dendritic cells in the spleen. Sci Immunol 2. jpet.aspetjournals.org

Lu R and Schmidtko A (2013) Direct intrathecal drug delivery in mice for detecting in vivo effects

of cGMP on pain processing. Methods Mol Biol 1020:215-221.

McDonald JG, Smith DD, Stiles AR and Russell DW (2012) A comprehensive method for at ASPET Journals on September 27, 2021

extraction and quantitative analysis of sterols and secosteroids from human plasma. J

Lipid Res 53:1399-1409.

Mutemberezi V, Buisseret B, Masquelier J, Guillemot-Legris O, Alhouayek M and Muccioli GG

(2018) Oxysterol levels and metabolism in the course of neuroinflammation: insights

from in vitro and in vivo models. J Neuroinflammation 15:74.

Mutemberezi V, Guillemot-Legris O and Muccioli GG (2016) Oxysterols: From cholesterol

metabolites to key mediators. Prog Lipid Res 64:152-169.

Pándy-Szekeres G, Munk C, Tsonkov TM, Mordalski S, Harpsøe K, Hauser AS, Bojarski AJ and

Gloriam DE (2017) GPCRdb in 2018: adding GPCR structure models and ligands.

Nucleic Acids Research 46:D440-D446.

Pereira JP, Kelly LM, Xu Y and Cyster JG (2009) EBI2 mediates B cell segregation between the

outer and centre follicle. Nature 460:1122-1126.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Preuss I, Ludwig MG, Baumgarten B, Bassilana F, Gessier F, Seuwen K and Sailer AW (2014)

Transcriptional regulation and functional characterization of the oxysterol/EBI2 system in

primary human macrophages. Biochem Biophys Res Commun 446:663-668.

Quinlan AR and Hall IM (2010) BEDTools: a flexible suite of utilities for comparing genomic

features. Bioinformatics 26:841-842.

Raithel SJ, Sapio MR, LaPaglia DM, Iadarola MJ and Mannes AJ (2018) Transcriptional

Changes in Dorsal Spinal Cord Persist after Surgical Incision Despite Preemptive Downloaded from Analgesia with Peripheral Resiniferatoxin. Anesthesiology 128:620-635.

Reboldi A, Dang EV, McDonald JG, Liang G, Russell DW and Cyster JG (2014) Inflammation.

25-Hydroxycholesterol suppresses interleukin-1-driven inflammation downstream of type jpet.aspetjournals.org

I interferon. Science 345:679-684.

Rosenkilde MM, Benned-Jensen T, Andersen H, Holst PJ, Kledal TN, Luttichau HR, Larsen JK,

Christensen JP and Schwartz TW (2006) Molecular pharmacological phenotyping of at ASPET Journals on September 27, 2021

EBI2. An orphan seven-transmembrane receptor with constitutive activity. J Biol Chem

281:13199-13208.

Rutkowska A, Dev KK and Sailer AW (2016a) The Role of the Oxysterol/EBI2 Pathway in the

Immune and Central Nervous Systems. Current Drug Targets 17:1851-1860.

Rutkowska A, O'Sullivan SA, Christen I, Zhang J, Sailer AW and Dev KK (2016b) The EBI2

signalling pathway plays a role in cellular crosstalk between astrocytes and

macrophages. Sci Rep 6:25520.

Rutkowska A, Preuss I, Gessier F, Sailer AW and Dev KK (2015) EBI2 regulates intracellular

signaling and migration in human astrocyte. Glia 63:341-351.

Rutkowska A, Sailer AW and Dev KK (2017) EBI2 receptor regulates myelin development and

inhibits LPC-induced demyelination. J Neuroinflammation 14:250.

Rutkowska A, Shimshek DR, Sailer AW and Dev KK (2018) EBI2 regulates pro-inflammatory

signalling and cytokine release in astrocytes. Neuropharmacology 133:121-128.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Smirnova NF, Conlon TM, Morrone C, Dorfmuller P, Humbert M, Stathopoulos GT, Umkehrer S,

Pfeiffer F, Yildirim AO and Eickelberg O (2019) Inhibition of B cell-dependent lymphoid

follicle formation prevents lymphocytic bronchiolitis after lung transplantation. JCI Insight

Sun S and Liu C (2015) 7alpha, 25-dihydroxycholesterol-mediated activation of EBI2 in immune

regulation and diseases. Front Pharmacol 6:60.

Team RC (2015) R: A Language and Environment for Statistical Computing in, R Foundation for Downloaded from Statistical Computing.

Toblin RL, Mack KA, Perveen G and Paulozzi LJ (2011) A population-based survey of chronic

pain and its treatment with prescription drugs. Pain 152:1249-1255. jpet.aspetjournals.org

Wanke F, Moos S, Croxford AL, Heinen AP, Graf S, Kalt B, Tischner D, Zhang J, Christen I,

Bruttger J, Yogev N, Tang Y, Zayoud M, Israel N, Karram K, Reissig S, Lacher SM,

Reichhold C, Mufazalov IA, Ben-Nun A, Kuhlmann T, Wettschureck N, Sailer AW, at ASPET Journals on September 27, 2021

Rajewsky K, Casola S, Waisman A and Kurschus FC (2017) EBI2 Is Highly Expressed

in Multiple Sclerosis Lesions and Promotes Early CNS Migration of Encephalitogenic

CD4 T Cells. Cell Rep 18:1270-1284.

Xing H, Chen M, Ling J, Tan W and Gu JG (2007) TRPM8 Mechanism of Cold Allodynia after

Chronic Nerve Injury. The Journal of Neuroscience 27:13680.

Yosten GLC, Harada CM, Haddock CJ, Giancotti LA, Kolar GR, Patel R, Guo C, Chen Z, Zhang

J, Doyle TM, Dickenson AH, Samson WK and Salvemini D (2020) GPR160 de-

orphanization reveals critical roles in neuropathic pain in rodents. The Journal of Clinical

Investigation 130: 2587-2592.

Zhuang ZY, Gerner P, Woolf CJ and Ji RR (2005) ERK is sequentially activated in neurons,

microglia, and astrocytes by spinal nerve ligation and contributes to mechanical

allodynia in this neuropathic pain model. Pain 114:149-159.

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

‡ Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021

‡ These studies were supported by the Saint Louis University start-up funds of Dr. Salvemini and Dr.

Arnatt; the National Institute of Health T32 Training Grant (GM008306-01; KB); and in part by National

Institute of Health National Heart, Lung, and Blood Institute (HL020948: JGM).

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

Figure 1: Gpr183 expression is upregulated in ipsilateral DH-SC from rats with CCI- induced neuropathic pain. A) Mechano-allodynia in male rats following CCI or sham surgery

(n=4/group). B) Gpr183 expression in the ipsilateral DH-SC from male rats on day 10 post CCI or sham surgery (n=4). Full image is found in Fig. S1. C) Quantification and D) representative image of Gpr183 expression in the spinal cord on day 10 CCI or sham surgery by RNAscope analyses. Image is stitched (x, y, and z; original magnification of 40x and axial depth of 6 μm) composite of spinal cord probed for Gfap (green), Gpr183 (red) and DAPI (blue). Signal was Downloaded from dilated by 3 pixel diameters for display purposes; box= area of quantification; ipsi=ipsilateral and contra=contralateral side to injury. Scale bar= 200 μm. Data are A) mean ± SD or C) median and analyzed by two-tailed, A) two-way repeated measures ANOVA (RM-ANOVA) with jpet.aspetjournals.org

Bonferroni comparisons or C) paired Student’s t test. *P<0.05 vs. Sham or # P<0.05 vs. Contra.

at ASPET Journals on September 27, 2021 Figure 2: Gpr183 expression is upregulated in microglia and astrocytes in ipsilateral DH-

SC from rats with CCI-induced neuropathic pain. A-C) Representative images (ipsilateral) with magnified inset to show detail and D-F) quantification of RNAscope signal for Gpr183

(yellow) and its colocalization with Aif1 (microglia; n=10; magenta; A,D), Gfap (astrocytes; n=10; magenta; B,E) or Rbfox3 (neurons; n=8; magenta; C,F) in superficial DH-SC (regions are marked by white box in Fig. 1D). Signal for lineage markers (magenta) was dilated by 0.5 pixel diameters to create a selection region that was applied to quantify Gpr183 for that cell lineage.

Scale bar= 50 μm. Data are median and analyzed by two-tailed, paired Student’s t test. *P<0.05 vs. Contra.

Figure 3: NIBR189 does not reverse CCI-induced mechanical allodynia. Acute intrathecal injections of NIBR189 (0.7μM, 2.3μM n=3; 23μM n=4) on D7 post CCI does not reverse

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

mechano-allodynia in male mice. Data are mean ± SD; Two-Way ANOVA with Dunnett’s multiple comparison; not significant p>0.05 vs D7.

Figure 4: Novel GPR183 antagonists inhibit GPR183-dependent 7α,25- dihydroxycholesterol induced calcium signaling in HL-60 cells. A) Workflow of in silico modeling; B) Structures, Enamine codes, and IC50’s of tested compounds C) Calcium mobilization dose response for HL-60 cells to 7α,25-OHC. Error bars represent mean ± SEM for

n=5. D) Dose-response of HL-60 cells to SAE-1, -10, and -14 inhibition of calcium mobilization Downloaded from induced by 7α,25-OHC (EC80 209 nM). Error bars represent mean ± SEM for n=4 E) Calcium mobilization dose response of HL-60 cells treated with or without Gpr183-targeting siRNA to

7α,25-OHC (n=3). F) Dose response of HL-60 cells treated with or without Gpr183-targeting jpet.aspetjournals.org siRNA to SAE-14 inhibition of calcium mobilization induced by 7α,25-OHC (EC80 209 nM). Data are mean ± SEM for n=3. at ASPET Journals on September 27, 2021

Figure 5: GPR183 antagonists reverse nerve injury-induced allodynia in mice. Acute intrathecal injections of SAE-1 (800nM), SAE-10 (1.4μM), or SAE-14 (2.9μM) were able to reverse CCI-induced mechano-allodynia on D7 post-surgery in male and female mice (data combined); Data are mean ± SD for n=4 per group; Two-Way ANOVA with Dunnett’s multiple comparison *p<0.05 vs D7.

Figure 6: Intrathecal administration of 7α,25-dihydroxycholesterol and its analog induces allodynia in mice. A) Acute intrathecal injection of 7α,25-OHC (24nM, n=6; 72nM, n=6 and

240nM, n=14) or its synthetic analog (SLUPP-1492; 615nM, n=8) induced mechano-allodynia in male and female mice (data combined) in a dose-dependent manner. B) Intrathecal 7α,25-OHC

(240nM) and SLUPP-1492 (615nM) also induced time-dependent development of cold-allodynia

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version.

(n=6/group). C,D) Pre-treatment with intrathecal SAE-14 (3µM mechano n=9/cold n=8; 1µM n=9; 0.3µM n=9), but not vehicle (mechano n=9/cold n=6) dose-dependently prevented the development of C) mechano- and D) cold-allodynia induced by 7α,25-OHC (480nM) in male and female mice. Data are mean ± SD and analyzed by two-way ANOVA with Dunnett’s comparison. *p<0.05 vs 0 hours; †p<0.05 vs Veh group. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021

JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021 JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021 JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021 JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021 JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021 JPET Fast Forward. Published on September 10, 2020 as DOI: 10.1124/jpet.120.000105 This article has not been copyedited and formatted. The final version may differ from this version. Downloaded from jpet.aspetjournals.org at ASPET Journals on September 27, 2021 GPR183-oxysterol axis in spinal cord contributes to neuropathic pain

K. Braden, L.A. Giancotti, Z. Chen, C. DeLeon, N. Latzo, T. Boehm, N. D’Cunha, B.M.

Thompson, T.M. Doyle, J.G. McDonald, J.K. Walker, G.R. Kolar, C.K. Arnatt, D. Salvemini

Journal of Pharmacology and Experimental Therapeutics

Supplementary Data and Methods

Supplementary Figure 1: Gpr183 expression the DH-SC from rats with CCI-induced neuropathic pain. Full representative image Fig. 1C showing Gpr183 expression in the dorsal and ventral spinal cord on day 10 post CCI or sham surgery by RNAscope analyses. Image is stitched (x, y, and z; original magnification of 40x and axial depth of 6 μm) composite of spinal cord probed for Gfap (green), Gpr183 (red) and DAPI (blue).

Supplemental Methods:

Custom FIJI Macro

Resultname = getTitle() setOption("Stack position", true);

for (n=1; n<=nSlices; n++) {

setSlice(n);

Stack.setChannel(2)

setThreshold(70, 255);

run("Create Selection");

run("Enlarge...", "enlarge=0.5 pixel");

Stack.setChannel(3);

setBackgroundColor(0, 0, 0);

run("Clear Outside", "slice");

run("Select None");

setThreshold(80, 255);

run("Analyze Particles...", "summarize slice");

}

Resultname = getTitle()

2 saveAs("Results", “filepath");

SLUPP-1492 SYNTHESIS:

Cholestene-7b-trifluoromethyl, 3b, 7a, 25-triol (SLUPP-1492) was prepared according to the following procedure.

25-Hydroxycholesterol 3β-acetate (SLUPP-1828)

To a solution of 25-hydroxycholesterol (1, 430 mg, 1.068 mmol), triethylamine (0.357 mL, 2.563 mmol) and acetic anhydride (0.122 mL, 1.281 mmol) in THF (10 mL) was added catalytic amount of 4-Dimethylaminopyridine (DMAP) (20-30mg) at room temperature. The reaction was stirred for 20 h and quenched with saturated solution of

NH4Cl in water and extracted with dichloromethane. The organic layer was combined and dried over Na2SO4. After filtration, the solvent was removed under reduced pressure and purified using flash chromatography (petroleum ether/ diethyl ether, 1:1) on silica gel. The residue was obtained as pure white powder 25-Hydroxycholesterol 3β-

1 acetate (2, 409 mg, 86%). H NMR (400 MHz, CDCl3) δ 5.37 (d, J = 4.9 Hz, 1H), 4.66 –

4.55 (m, 1H), 2.32 (d, J = 7.2 Hz, 2H), 2.03 (s, 3H), 2.01 – 1.92 (m, 2H), 1.90 – 1.78 (m,

3H), 1.57 (s, 6H), 1.52 – 1.34 (m, 8H), 1.21 (s, 8H), 1.18 – 1.04 (m, 5H), 1.02 (s, 4H),

0.93 (d, J = 6.5 Hz, 3H), 0.68 (s, 3H). LC-MS: found [(M-AcOH)+] m/z = 385.3, expected

(m/z = 385.3)

(3β)-3-(Acetyloxy)-25-hydroxycholest-5-en-7-one (SLUPP-1829)

To a solution of 25-Hydroxycholesterol 3β-acetate (2, 409 mg, 0.9 mmol) in ethyl acetate (15 mL) and molecular sieves (4A) was added t-butylhydroperoxide (0.8 mL, 5 mmol) and stirred under nitrogen for 20 min. To the reaction mixture, manganese acetate (25 mg, 0.09 mmol) was added and stirred at room temperature for 48 h. The solids were filtered off through a pad of celite. The solvent was removed under reduced pressure and purified using flash Chromatography (petroleum ether/ diethyl ether, 1:2) on silica gel. The residue was obtained pure white powder (3β)-3-(Acetyloxy)-25-

1 hydroxycholest-5-en-7-one (3, 248 mg, 59%). H NMR (400 MHz, C6D6) δ 5.73 (d, J =

1.7 Hz, 1H), 4.75 – 4.65 (m, 1H), 3.04 – 2.93 (m, 1H), 2.34 – 2.26 (m, 1H), 2.16 (s, 1H),

2.08 – 1.90 (m, 3H), 1.83 – 1.75 (m, 1H), 1.73 (s, 3H), 1.54 – 1.28 (m, 11H), 1.26 – 1.17

(m, 3H), 1.15 – 1.05 (m, 2H), 1.10 (s, 6H), 1.02 (d, J = 6.5 Hz, 3H), 0.99 – 0.94 (m, 1H),

0.74 (s, 3H), 0.71 – 0.65 (m, 1H), 0.61 (s, 3H). ). LC-MS: found [(M-AcOH)+] m/z =

399.3, expected (m/z = 399.3)

Cholestene-7b-trifluoromethyl, 3b, 7a, 25-triol (SLUPP-1492, 5)

To a solution of the above acetate (0.247 g, 0.539 mmol) in DME (2.0 mL) was added cesium fluoride (8 mg, 0.054 mmol) and then a DME (0.25 mL) solution of trifluoromethyl trimethylsilane (0.169 g, 1.19 mmol)). The resulting mixture was stirred for 2 days at room temperature after which the reaction was concentrated to dryness.

The resulting residue was taken up in 1% EtOAc-hexanes and subjected to flash chromatography, Elution with EtOAc-Hexanes (1% to 5% to 20%) led to isolation (0.191 g, 0.36 mmol) of a glassy solid that appears to be 25-trimethylsiloxy starting material.

This material was re-subjected to the trifluoromethylation conditions as described below.

The material isolated from above (0.191 g, 0.36 mmol) was dissolved in DME (1.0 mL) and stirred at room temperature. To this mixture was added cesium fluoride (5 mg, 0.04 mmol) followed by addition of a DME (0.5 mL) solution of trifluoromethyl trimethylsilane

(0.113 g, 0.792 mmol). The reaction was stirred at room temperature for 24h. Some starting material remained so additional trifluoromethyl trimethylsilane (0.051 g, 0.360 mmol) was added and the reaction and was stirred for another 24h. The reaction was judged complete by thin-layer chromatography (TLC) and the reaction was concentrated to dryness. The resulting residue was dissolved in 1% EtOAc-Hexanes and purified by flash chromatography. Elution with EtOAc-hexanes (1% to 10%) led to the isolation

(0.223 mg) of a complex mixture of mono- and bis-silylated products with incorporation of the trifluoromethyl group.

The above mixture was dissolved in MeOH (3 mL) and to this was added potassium carbonate (0.154, 1.1 mmol) and the mixture was stirred at room temperature. The solution turned pale yellow upon addition of the potassium carbonate. The reaction was allowed to stir overnight at room temperature. The reaction was judged complete by

TLC and the reaction was filtered through a small pad of celite washing with MeOH and then EtOAc. The organics were combined and concentrated. The residue was dissolved in small amount of MeOH and absorbed onto silica and purified by flash chromatography. Elution with EtOAc-hexanes (5-50%) led to isolation of the desired product (0.066 g) as mixture of diastereomers. The material was dried overnight under vacuum in a 1-dram vial. The resulting residue was dissolved in hot EtOAc and the vial was then allowed to cool to room temperature. After ~ 1h a precipitate resulted. The

EtOAc was carefully removed and the resulting solid was washed with cold EtOAc and filtered to give the desired product (0.025 g) in > 90% purity as a white crystalline solid.

1 H NMR (400 MHz, CDCl3) δ 5.31 (s, 1H), 3.55 (s, 1H), 2.48 – 2.37 (m, 1H), 2.30 – 2.19

(m, 1H), 2.04 – 1.84 (m, 5H), 1.67 – 1.58 (m, 1H), 1.56 – 1.47 (m, 5H), 1.46 – 1.32 (m,

7H), 1.29 – 1.15 (m, 11H), 1.15 – 1.01 (m, 5H), 0.98 – 0.89 (m, 6H), 0.72 (s, 3H). LC-

MS, found [M + Na] m/z = 509.3 (expected m/z = 509.3) and [(M + H) - 2H20] m/z =

451.3 (expected 451.3)

REACTION SCHEME

Spectral Data for compounds 2, 3 and 5 (SLUPP1492)