Fc gamma receptor IIa signaling from spinal cord astrocytes promotes neuropathic pain in rats

Abstract

Injury to somatosensory nerves can lead to neuropathic pain. We recently identified that B cells play a crucial role in the development of neuropathic pain through a mechanism involving secreted immunoglobulin G (IgG) signaling at Fc gamma receptors (FcγRs). Here, we demonstrate that Fc gamma receptor IIa (FcγRIIa), expressed by astrocytes in the spinal cord, contributes to the development of mechanical allodynia after nerve injury in male and female rats. Following unilateral chronic constriction injury (CCI) of the sciatic nerve, Fcgr2a gene transcription increased specifically in the ipsilateral dorsal horn of the spinal cord, but remained unaltered in the dorsal root ganglia (DRGs) and contralateral spinal cord. FcγRIIa immunoreactivity increased in the ipsilateral spinal dorsal horn after CCI, and its expression colocalized primarily with GFAP⁺ astrocytes. Genetic disruption of Fcgr2a in GFAP-expressing spinal astrocytes using adeno-associated virus (AAV)-mediated CRISPR-Cas9 gene editing attenuated mechanical allodynia for weeks after CCI. In purified cultures of primary astrocytes, IgG immune complexes (IgG-IC) increased transcription of proinflammatory cytokines and chemokines. Expression of these cytokines and chemokines was reduced by siRNA-mediated knockdown of Fcgr2a, or by inhibition of the FcγRIIa effectors spleen tyrosine kinase (Syk) or nuclear factor-κB (NF-κB). These data suggest that FcγRIIa expressed by spinal astrocytes are activated following peripheral nerve injury and may directly contribute to injury-induced tactile pain through the release of proinflammatory mediators. These findings expand our understanding of how neuroimmune signaling from astrocytes contributes to the development of mechanical allodynia.

Introduction

Chronic pain impacts approximately 20% of adults in the United States and represents an ongoing health crisis of diminished physical and emotional well-being^1,2. A common and debilitating form of chronic pain is neuropathic pain, which arises from damage or disease of the somatosensory nervous system³. Mechanical allodynia, in which normally innocuous light touch is perceived as painful, is a canonical symptom of neuropathic pain that is particularly recalcitrant to opioids or other pain medications^4,5. Uncovering the biological substrates of allodynia is therefore needed to develop new, safe, and efficacious therapeutics for neuropathic pain.

We recently demonstrated that B cells play a critical role in the development of neuropathic pain in mice via secreted IgG^6–8. After chronic constriction injury (CCI), autoreactive IgG is found in mouse serum, and IgG accumulates in the dorsal root ganglia (DRG) and spinal cord of mice and humans. Antigen-IgG immune complexes form, which can activate Fc gamma receptors (FcγRs) to promote nociceptor hyperexcitability and mechanical allodynia^6,8. CCI-induced allodynia is reduced in rodents with global genetic disruption of FcγR signaling or conditional deletion of Fcgr1 in DRG neurons^6,9. However, the role of other FcγR subtypes in neuropathic pain, as well as their action beyond the DRG is still largely unknown. Four different FcγR subtypes are described in rats (Fcgr1a, Fcgr2a, Fcgr2b, and Fcgr3a), which are conserved in humans^10,11. While these subtypes differ in their structure and antibody affinity, they share the characteristic of being activated by IgG that has bound its antigen and aggregated into an IgG immune complex (IgG-IC). Activation of these receptors on immune cells induces a range of effector functions, including cytokine release^12,13. In an unbiased transcriptomic analysis of the spinal cord dorsal horn after CCI, the intensity of allodynia was strongly correlated with transcriptional pathways associated with FcγRs¹⁴. Among these receptors, Fcgr2a (the gene encoding CD32a, referred hereafter as FcγRIIa) had the strongest association with increased allodynia¹⁴. However, the functional significance of this FcγR subtype in pain remains to be determined.

In this study, we sought to characterize FcγRIIa expression in the spinal cord, investigating the hypothesis that FcγRIIa contributes to mechanical allodynia after neuropathic injury. We demonstrate that FcγRIIa is expressed by spinal cord astrocytes in rats, and that activation of this receptor promotes release of proinflammatory and pronociceptive cytokines. These findings expand our understanding of how IgG accumulation in the lumbar spinal cord contributes to allodynia after peripheral nerve injury.

Methods

Animals

Male and female Sprague Dawley rats were used as subjects for all experiments. Adult rats (8–10 weeks old) were ordered from Inotiv and given at least 1 week to acclimate to the pathogen-free, AAALAC-accredited vivarium at the University of Texas MD Anderson Cancer Center before any manipulations were performed. Breeding was performed in-house only to generate subjects for astrocyte immunopanning experiments. Rats were housed 2–3 per cage in individually ventilated cages (10.5 in × 16.75 in × 7.25 in; Tecniplast) with ad libitum access to drinking water and food (PicoLab Rodent Diet 5053; Purina). Each cage was lined with absorbable corncob bedding and contained shredded paper nesting material and wooden gnawing blocks for enrichment, and rats were transferred to clean cages once per week. The colony was maintained in temperature- and humidity-controlled rooms on a 12-hour light-dark cycle (lights on from 7:00 AM to 7:00 PM), and procedures were performed during the light cycle. Sample sizes were not determined by statistical methods but based on prior studies using similar approaches^6,9,15–17. Animals were randomly assigned to treatment or control groups using an online group randomization tool (GraphPad Software). All animal experiments were performed in accordance with an approved protocol by the Institutional Animal Care and Use Committee at the University of Texas MD Anderson Cancer Center.

Chronic constriction injury (CCI) surgery

Neuropathic pain was modeled using unilateral CCI of the sciatic nerve¹⁸ as we described previously^19,20. Surgeries were performed with aseptic technique using autoclaved surgical instruments that were heated in glass bead sterilizers between surgeries. Rats were anesthetized via inhalation of isoflurane (2–3% in oxygen) and placed on an electric heating pad to maintain body temperature. Ophthalmic ointment was applied to each eye to prevent drying. Fur around the left leg was shaved with an electric razor, and the skin around the surgical site was cleansed with povidone-iodine and 70% ethanol. The skin was incised using a scalpel blade, and the left sciatic nerve proximal to the nerve trifurcation was exposed through blunt dissection of the biceps femoris muscle. Using glass nerve hooks constructed from flame-polished Pasteur pipettes, a segment of the sciatic nerve was gently liberated from the surrounding connective tissue and externalized. For CCI surgeries, 4 sutures (4–0 chromic gut; Ethicon) were tied loosely around the sciatic nerve; each ligature was fasted with a double knot and placed approximately 1 mm apart from each other along the nerve. For sham surgeries, the sciatic nerve was isolated and manipulated in an identical fashion, but no chromic gut sutures were applied. The muscle layer was sutured closed (4–0 silk; Ethicon), and the skin was closed with 9 mm wound clips. At the completion of surgery, rats were returned to their home cage and monitored postoperatively until fully ambulatory. For experiments with naive subjects, rats were left undisturbed in their home cages.

Behavioral tests for mechanical allodynia

In advance of any behavioral measurements, rats were handled for 5 minutes per day over 3 days to familiarize subjects to the investigator. Rats were habituated to the testing apparatus in the presence of the investigator for 60 minutes per day on 3 separate days prior to behavioral testing. The testing apparatus for measuring mechanical allodynia consisted of acrylic enclosures placed on a raised metal mesh platform. On testing days, rats were placed within their acrylic enclosures and allowed to acclimate for 30 minutes before testing. Measurements were taken prior to CCI or sham surgery (baseline), at 3- or 4-day intervals during the first 2 weeks after surgery, and at 7-day intervals thereafter, with measurements continuing for 8 weeks after surgery. Male and female rats were tested during separate sessions, and the entire testing apparatus was cleaned with Peroxigard between sessions to disinfect and deodorize surfaces. The investigator was blinded to group allocation when performing behavioral measurements.

von Frey test for punctate allodynia

The von Frey test was conducted to measure punctate mechanical allodynia as previously described¹⁵. A series of 10 Semmes-Weinstein monofilaments (Stoelting) with a calibrated bending force of 0.4, 0.6, 1, 1.4, 2, 4, 6, 8, 10, and 15 g were used. Filaments were applied perpendicularly to the plantar surface of the hindpaw and pressed until bending of the filament was observed. A positive response was characterized by a rapid withdrawal of the hindpaw away from the filament within 3 seconds of stimulation. The threshold stimulus intensity required to elicit a paw withdrawal response was determined based on the up-down method²¹. The 2 g filament was first applied 3 times to each hindpaw. If 0 or 1 response was recorded, the next higher filament (4 g) was applied, continuing upward until 3 positive responses were recorded; if 2 or 3 positive responses were recorded at 2 g, the lowest filament (0.4 g) was tested, gradually increasing to higher filaments until 3 positive responses were recorded. The paw withdrawal responses were used to calculate absolute threshold (the 50% probability of response) by fitting a Gaussian integral psychometric function using a maximum-likelihood fitting method^22,23 using PsychoFit software (v. 1.3.1) as previously described²⁴. Estimated thresholds derived from a Gaussian integral function yield a mathematical continuum that are appropriate for parametric statistical analyses.

CRISPR-Cas9 gene editing of Fcgr2a in spinal cord astrocytes

Adeno-associated virus (AAV) constructs

AAV constructs were designed to deliver under the control of the GFAP promoter either Cas9 from Staphylococcus aureus (saCas9, referred hereafter as Cas9) for gene editing or green fluorescent protein (GFP) as a control, along with single guide RNA (sgRNA) for Fcgr2a. CRISPR-Cas9 gene editing plasmid constructs were designed and grown as described²⁵. The AAV-GFAP-Cas9 backbone with HA-tag was obtained commercially (#7125; Vector Biolabs). The sgRNA was designed by Genecopoeia to target Fcgr2a and oligonucleotides were synthesized by IDT (guide sequences, 5’->3’: top: CACCGTCCTTCCAGAAACACCAGTTG; bottom: AAACCAACTGGTGTTTCTGGAAGGAC). For the control vector, Cas9 was replaced with GFP which was amplified from the plasmid (#82416; Addgene). All the vector sequences were confirmed by DNA sequencing. Production of AAVs (using the AAV5 serotype) was performed by the Stanford Gene Vector and Virus Core (RRID: SCR_023250).

AAV injections

To selectively edit Fcgr2a in spinal cord astrocytes, rats were given injections of AAV5-GFAP-Fcgr2a-Cas9 or AAV5-GFAP-Fcgr2a-GFP into the spinal dorsal horn. Intraspinal dorsal horn viral injections without laminectomy were performed as previously described²⁶. Rats were anesthetized via inhalation of isoflurane (2–3% in oxygen). Fur on the dorsal surface along the vertebral column was shaved with an electric razor, and the skin was cleansed with povidone-iodine and 70% ethanol. The skin along the spine from T12-L3 was incised and separated, and muscles on both sides of the vertebral column between L1-L2 vertebrae were incised. A vertebrae clamp attached to a spinal frame (Kopf Instruments) was inserted between L1-L2 to elevate and support the vertebral column. Paraspinal muscles over T13-L1 vertebrae were removed, and a 30G needle was used to incise a small window into the meningeal layers. The AAV was loaded into a NanoFil syringe with 35G needle (World Precision Instruments) fitted on a UMP3 microinjection syringe pump (World Precision Instruments). The needle was inserted into the superficial dorsal horn and the viral injection (800 nL, 8 × 10⁹ vg) was infused over 60 seconds. The needle was left in place for several minutes after infusion and then slowly withdrawn. The skin was stapled closed with 9 mm wound clips, and rats were administered meloxicam (4 mg/kg, s.c.). Rats were given 18 days for recovery and viral transduction before CCI surgeries were performed.

RNA isolation, RT-qPCR, and RT-PCR

Rats were euthanized with an injection of pentobarbital-phenytoin solution (i.p.) and transcardially perfused with 100 mL of ice-cold 0.9% saline. Lumbar L4–5 spinal cord segments and L4–5 DRGs were extracted following laminectomy. Spinal cord segments were dissected on an ice-cold petri dish to isolate the ipsilateral and contralateral dorsal quadrants. All samples were placed in 1.5 mL microcentrifuge tubes, snap-frozen through submersion in liquid nitrogen, and stored at −80 °C. RNA was isolated by mechanically dissociating tissues in TRIzol (Thermo Fisher Scientific), performing phase separation with chloroform (Sigma-Aldrich), and precipitating in isopropanol (Sigma-Aldrich) with glycogen (Thermo Fisher Scientific). RNA was washed twice with 75% ethanol, resuspended in nuclease-free water (Thermo Fisher Scientific), and stored at −80 °C. RNA concentration was determined using a BioTek Synergy HTX spectrophotometer (Agilent Technologies). cDNA was generated by reverse transcription of RNA using iScript gDNA Clear kits (Bio-Rad). Reverse transcription quantitative PCR (RT-qPCR) was performed by diluting cDNA and associated primers (see Table S1 for primer sequences) in SsoAdvanced Universal SYBR Green Supermix (Bio-Rad) and monitoring the reaction with the CFX Connect Real-Time PCR Detection System (Bio-Rad). Relative gene expression against Gapdh using the 2^−ΔΔCT method was performed as previously described²⁷. RT-PCR was performed by diluting cDNA and associated primers in MyTaq HS Red Mix (Meridian Bioscience) and visualizing the amplified product by agarose gel electrophoresis.

Immunohistochemistry

Perfusion and tissue preparation

Rats were euthanized with an injection of pentobarbital-phenytoin solution (i.p.) and transcardially perfused with 100 mL of ice-cold 0.9% saline, followed by 100 mL of ice-cold 4% paraformaldehyde. Spinal cords were extracted following laminectomy, and tissues were post-fixed in 4% paraformaldehyde overnight at 4 °C. Samples were cryoprotected stepwise in 15%, 22%, and 30% sucrose in 0.1 M phosphate buffer (PB) with 0.01% sodium azide at 4 °C. Spinal cord lumbar enlargements were dissected, embedded in O.C.T compound (Sakura Finetek), and frozen for cryosectioning. Tissues were serially sectioned at 16 μm using a cryostat (Leica Biosystems), thaw mounted to SuperFrost Plus charged slides (Thermo Fisher Scientific), and stored at −20 °C.

Immunofluorescence

Slides were removed from −20 °C storage, and a hydrophobic barrier was drawn around the tissue samples with an ImmEdge PAP pen (Vector Laboratories). Slides were washed (0.3% Triton X-100 (Sigma-Aldrich) in PBS; 10 minutes × 2 washes) and incubated with blocking buffer (5% normal donkey serum (NDS; Abcam), and 0.3% Triton X-100 in PBS) for 1 hour at room temperature in a light-protected slide staining tray. Slides were washed (10 minutes × 1 wash) and incubated with primary antibodies diluted in antibody buffer (2% NDS and 0.3% Triton X-100 in PBS) overnight at 4 °C. Subsequently, sections were washed (10 minutes × 3 washes) and incubated with secondary antibodies diluted in antibody buffer (2% NDS and 0.3% Triton X-100 in PBS) for 2 hours at room temperature. Slides were then washed (10 minutes × 3 washes), incubated with 4′,6-diamidino-2-phenylindole (DAPI) (1:5,000; Sigma-Aldrich) nuclear counterstain for 5 minutes at room temperature, washed in PBS (10 minutes × 4 washes), cover slipped with FluorSave mounting medium (Millipore Sigma), and stored at 4 °C. As a control, slides were processed concurrently with no primary antibody included in the antibody buffer. Primary and secondary antibodies and their dilutions are provided in Table S2.

Widefield image acquisition

To verify that intraspinal injections led to viral recombination in the ipsilateral dorsal horn, widefield fluorescent micrographs were captured using an Olympus BX53 fluorescence microscope with DP80 digital camera (Olympus) at 4x magnification. Inclusion criteria for subsequent analysis of behavioral data was identification of GFP or HA-tag (for Cas9 virus) within the ipsilateral dorsal horn.

Confocal image acquisition and analysis

To assess FcγRIIa immunofluorescence in spinal cord, fluorescent photomicrographs were captured using a Nikon Eclipse Ti2 confocal microscope (Nikon Instruments) at 20x magnification. Z-stacks were acquired for the entire thickness of each tissue section using optimal settings and step size (0.9 μm) set by Nikon NIS-Elements software (Nikon Instruments). Maximum intensity projections of each channel were analyzed using Fiji/ImageJ software (v. 1.53t). Regions of interest were manually drawn around the spinal cord dorsal horn (excluding dorsal columns and areas of the image outside the spinal cord), and the mean gray intensity of FcγRIIa was measured for each animal. Two to 3 sections were analyzed per animal, and the average mean gray intensity was calculated for each animal. For FcγRIIa colocalization, fluorescent photomicrograph Z-stacks were captured using a Nikon Eclipse Ti2 confocal microscope (Nikon Instruments) at 40x magnification. Z-stacks were acquired for the entire thickness of each tissue section using optimal settings and step size (0.2 μm) set by Nikon NIS-Elements software. 3D projections of FcγRIIa and cell-specific markers (GFAP, IBA1, and MAP2) or HA-tag and GFAP were analyzed using Just Another Colocalization Plugin (JACoP, ImageJ). Regions of interest were manually drawn (excluding dorsal columns and areas of the image outside the spinal cord) and the default JACoP plugin settings were used to calculate the Manders’ coefficients above threshold (tM) for FcγRIIa relative to the either GFAP, IBA1 or MAP2, and HA-tag relative to GFAP for each image. Two sections were analyzed per animal, and the average Manders’ coefficient above threshold was calculated for each animal. Analysis was performed by an experimenter blinded to the experimental group of each image.

Astrocyte immunopanning and in vitro experiments

Immunopanning

Astrocytes were purified and cultured from rat forebrains between postnatal days 6 and 8 (P6–8) by immunopanning, based on previously published protocols^16,28. Whole cortex was blunt dissected from rat brains in Dulbecco’s PBS (D-PBS; Cytiva), and cortices from both male and female rats were combined for further processing. After removing meninges with fine forceps under a dissection microscope, the cortices were enzymatically digested with papain in a 6 cm Petri dish for 40 minutes at 34 °C and 10% CO₂. Tissues were dissociated by mechanical trituration with a 5 mL serological pipette and filtered through a 20 μm Nitex mesh filter to generate a single cell suspension. Cells were then passed through a series of antibody- or lectin-coated 15 cm Petri dishes to trap and deplete endothelial cells (Griffonia simplicifolia lectin, BSL-1; Vector Laboratories), microglia and macrophages (CD45 antibody; BD Biosciences), and oligodendrocytes (O4 hybridoma)²⁹. Astrocytes were then positively selected by panning cells on an antibody-coated 15 cm Petri dish (ITGB5; eBioscience). Astrocytes were detached from the dish with 0.025% trypsin (Sigma-Aldrich) in Earle’s balanced salt solution (EBSS; Sigma-Aldrich) and dislodged by repeated aspiration and dispensing of 20 mL of 30% fetal bovine serum (FBS; Gibco). Purified astrocytes were centrifuged at 170 rcf for 11 minutes at room temperature, resuspended in culture media, and counted by trypan blue exclusion. Astrocytes were then plated at 25,000 cells per well in 24 well cell culture plates coated with poly-D-lysine (Gibco). Astrocytes were cultured in a serum-free medium of 50% Dulbecco’s modified Eagle’s medium (DMEM; Gibco) and 50% Neurobasal medium (Gibco) supplemented with penicillin-streptomycin (100 U/mL penicillin, 100 μg/mL streptomycin; Gibco), sodium pyruvate (1 mM; Gibco), L-glutamine (292 μg/mL; Gibco), SATO supplement (containing BSA (100 μg/mL; Sigma-Aldrich), transferrin (100 μg/mL; Sigma-Aldrich), putrescine dihydrochloride (16 μg/mL; Sigma-Aldrich), progesterone (60 ng/mL; Sigma-Aldrich), sodium selenite (40 ng/mL; Sigma-Aldrich)), N-acetylcysteine (5 μg/mL; Sigma-Aldrich), and heparin-binding EGF-like growth factor (HBEGF, 5 ng/mL; Sigma-Aldrich). Astrocyte cultures were maintained in a humidified incubator at 37 °C and 10% CO₂, and half-media exchanges with fresh media were performed every 2–3 days. Cells were visualized in each well using phase contrast imaging on an EVOS microscope (Thermo Fisher Scientific).

IgG-IC, siRNA, and drug treatments

After 5 days in vitro (DIV5), astrocytes were transfected with Fcgr2a siRNA or negative control siRNA (15 pmol, Silencer Select; Thermo Fisher Scientific) using Lipofectamine 3000 transfection reagent (Thermo Fisher Scientific), fully replacing media after 24 hours. On DIV7, FcγRs on cultured astrocytes were stimulated with IgG-IC. IgG-IC was formed by combining purified mouse IgG F(ab’)₂ fragment (as antigen) and rat anti-mouse IgG (as antibody) in a ratio of 1:1 by volume (after diluting in sterile PBS to equivalent concentrations) and incubating in a 25 °C oven for 1 hour, based on previous methods³⁰. Dilutions of IgG-IC (3–10 μg/mL) were prepared in culture media and dispensed into wells, with media treatment used as a reference control. Following a 4-hour incubation, the culture media was aspirated, cells were lysed with TRIzol, and samples were processed for RT-qPCR; alternatively, following a 24-hour incubation, the culture media were aspirated to quantify secreted cytokines. R406 (Cayman Chemical) and ammonium pyrrolidinedithiocarbamate (PDTC; Sigma-Aldrich) were dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich) and prepared by diluting in culture media. Wells were pretreated with R406 (3 μM) or PDTC (30 μM) 1 hour before IgG-IC treatment. Alternatively, DIV7 astrocytes were either fixed with 4% paraformaldehyde and processed for immunofluorescence against GFAP (using the same protocol listed here, but with normal goat serum (NGS; Abcam) replacing NDS), or collected for RT-PCR using cell-specific markers (versus whole brain as positive control) based on previous reports¹⁶.

ELISA

Astrocyte culture media were collected 24 hours following IgG-IC treatment. Supernatant levels of interleukin-1 beta (IL-1β) and tumor necrosis factor (TNF) were analyzed using Quantikine enzyme-linked immunosorbent assay (ELISA) kits for rat IL-1β or TNF (R&D Systems). ELISAs were performed following manufacturer’s instructions, with 50 μL culture supernatant loaded per sample. Plates were read on a BioTek Synergy HTX spectrophotometer (Agilent Technologies) to measure optical density and calculate protein concentration. All samples and standards were run in duplicate, and concentration of samples that were within the linear range was quantified based on the standard curve and normalized to pg/mL.

Statistics

Data are presented as group mean ± SEM, with individual subjects depicted as symbols whenever feasible. Data from both males and females were combined to form a single group, and group sizes are noted in each figure legend. Statistical analyses and figure generation were performed with GraphPad Prism software (v. 10.4.2). Data from von Frey testing were analyzed by repeated measures two-way ANOVA and Šídák’s post hoc tests. Immunofluorescence data were analyzed by unpaired t test or two-way ANOVA and Šídák’s post hoc test. Data from RT-qPCR and ELISA were analyzed by unpaired t test or Kruskal-Wallis and Dunn’s post hoc test. Complete details of statistical methods, comparisons, and results are presented in Table S3. Data exclusion criteria involved quality control exclusion of RT-qPCR samples if housekeeping gene amplification varied more than 2 cycles from the group mean (5 samples total excluded). No method for outlier detection was performed, and no other data points were excluded from analyses. Differences between groups were considered statistically significant when P < 0.05.

Results

Peripheral nerve injury upregulates Fcgr2a selectively in the ipsilateral L4–5 spinal dorsal horn

Prior work analyzing the spinal cord transcriptome after CCI revealed a correlation between the severity of mechanical allodynia and Fcgr2a expression in the ipsilateral dorsal spinal cord at a single timepoint in male rats¹⁴. Here, we sought to characterize gene transcription of Fcgr2a in male and female rats at multiple timepoints after nerve injury in the ipsilateral and contralateral L4–5 dorsal spinal cord and their associated L4–5 DRGs. Compared to naive controls (day 0), we observed a significant increase in Fcgr2a mRNA expression in the ipsilateral L4–5 dorsal spinal cord after CCI that peaked at day 10 and day 14 after injury, returning to baseline expression by day 70 (Figure 1A, S1A). These timepoints of heightened Fcgr2a expression correspond with the onset of maximal allodynia in the rat CCI model, while the reduced expression coincides with resolution of CCI-induced allodynia at around day 70¹⁹. We found no change in Fcgr2a mRNA expression in the contralateral spinal cord from the same animals at any timepoint from day 3 to day 70 after CCI (Figure 1B, S1B). Likewise, there were no significant changes in Fcgr2a expression in the ipsilateral L4–5 DRGs at any timepoint after CCI compared to controls (Figure 1C, S1C), nor were any changes in Fcgr2a expression observed in the contralateral L4–5 DRGs (Figure 1D, S1D). We also evaluated expression in the ipsilateral cervical C3–4 segment of the spinal cord (which receives no innervation from the sciatic nerve) and found no differences in Fcgr2a expression at any timepoint after CCI (Figure 1E, S1E). These findings suggest that sciatic nerve injury induces dynamic upregulation of Fcgr2a localized exclusively within the ipsilateral L4–5 spinal cord at the site of sciatic innervation.

Figure 1. Fcgr2a expression increases in ipsilateral L4–5 spinal cord after peripheral nerve injury.

(A to E) Fcgr2a relative gene expression was quantified with RT-qPCR from tissues collected after unilateral chronic constriction injury (CCI) of the sciatic nerve. Tissues were collected from naive control rats (day 0) and from nerve-injured animals at days 3, 7, 10, 14, 21, 35, and 70 after CCI. Fcgr2a was measured in (A) ipsilateral L4–5 spinal cord dorsal horn, (B) contralateral L4–5 spinal cord dorsal horn, (C) ipsilateral L4–5 DRGs, (D) contralateral L4–5 DRGs, and (E) ipsilateral C3–4 spinal cord dorsal horn. n = 5–8 rats per group (2–5 males and 2–3 females). Kruskal-Wallis and Dunn’s post hoc test; *P < 0.05, **P < 0.01 compared to day 0. Data are presented as mean ± SEM.

FcγRIIa increases in the spinal dorsal horn after nerve injury and is primarily expressed by GFAP⁺ astrocytes

To validate our gene expression results, we examined immunoreactivity of FcγRIIa in the ipsilateral L4–5 spinal cord dorsal horn on day 14 after peripheral nerve injury with confocal microscopy (Figure 2A). FcγRIIa immunoreactivity increased in the spinal cord of animals that received CCI surgery compared to sham controls (Figure 2B). To determine the cellular expression of FcγRIIa within the dorsal horn, we labeled spinal cord sections for FcγRIIa and analyzed the extent of colocalization with either GFAP⁺ astrocytes, IBA1⁺ microglia, or MAP2⁺ neurons (Figure 2C). We measured colocalization with Manders’ coefficient, which applies a binary threshold to each fluorescence channel and calculates the incidence of overlap for FcγRIIa with each of the cell-specific markers. Quantification revealed that FcγRIIa⁺ signal overlapped with GFAP⁺ astrocytes most prominently, with greater colocalization in astrocytes compared to microglia or neurons (Figure 2D). No differences were observed in the Manders’ coefficient values between sham or CCI animals, suggesting that FcγRIIa signal above background is most likely to colocalize with GFAP⁺ astrocytes regardless of injury status (Figure 2D). These data indicate that the intensity of FcγRIIa immunofluorescence increases in the dorsal horn after nerve injury, and the expression of FcγRIIa in the spinal cord is localized primarily on GFAP⁺ astrocytes.

Figure 2. FcγRIIa increases expression and colocalizes with GFAP⁺ astrocytes in ipsilateral spinal cord dorsal horn after nerve injury.

(A) Representative 20x images of ipsilateral L4–5 spinal cord dorsal horn 14 days after sham or CCI surgery. Immunofluorescence depicts FcγRIIa (red) and DAPI (blue). Scale bars indicate 100 μm. (B) Mean gray intensity for FcγRIIa was quantified in ipsilateral dorsal horn 14 days after sham or CCI surgery. n = 8 male rats per group. Unpaired t test; ***P < 0.001. (C) Representative 40x images of ipsilateral L4–5 spinal cord dorsal horn 14 days after sham or CCI surgery. Scale bars indicate 100 μm. Insets display magnified images of highlighted regions (dashed white boxes, 100 μm × 100 μm area). Immunofluorescence depicts FcγRIIa (red) and markers for astrocytes (GFAP), microglia (IBA1), and neurons (MAP2) (each in green). (D) Colocalization of FcγRIIa with GFAP, IBA1, and MAP2 was quantified with Manders’ coefficient (tM) in ipsilateral dorsal horn 14 days after sham or CCI surgery. n = 8 male rats per group. Two-way ANOVA and Šídák’s post hoc test; ***P < 0.001 compared to IBA1; ^^^P < 0.001 compared to MAP2. Data are presented as mean ± SEM.

Fcgr2a contributes to mechanical allodynia after peripheral nerve injury

To test whether there is a functional role for FcγRIIa in spinal cord astrocytes in the development of neuropathic pain, we deployed a cell type-specific CRISPR-Cas9 gene editing approach to knockdown Fcgr2a in spinal cord astrocytes. To accomplish this, we performed intraspinal injections of AAV5-GFAP-Fcgr2a-Cas9 or AAV5-GFAP-Fcgr2a-GFP control, and 18 days later, animals received unilateral CCI of the sciatic nerve. Rats injected with Cas9 virus to knockdown Fcgr2a exhibited attenuated mechanical allodynia to innocuous punctate stimuli compared to rats that received control GFP virus, with a significant main effect of Fcgr2a knockdown (Figure 3A, S2A). Withdrawal thresholds in the contralateral uninjured paw were no different between groups at baseline or after CCI for punctate stimuli (Figure 3B, S2B), suggesting no effect of virus treatment on nociceptive thresholds in the absence of injury. Using immunofluorescence, we verified viral transduction of astrocytes in the ipsilateral dorsal horn by staining for HA-tag (to visualize Cas9 viral expression) alongside GFAP (Figure 3C). Colocalization analysis revealed overlap of HA-tag with GFAP, suggestive of transduction of spinal astrocytes (Figure 3D). Collectively, these results support the notion that Fcgr2a in GFAP⁺ astrocytes in the spinal cord contributes to the development of mechanical hypersensitivity following peripheral nerve injury.

Figure 3.

CRISPR-Cas9 editing of Fcgr2a in spinal cord GFAP⁺ astrocytes in vivo attenuates mechanical allodynia after nerve injury. (A) von Frey thresholds for punctate mechanical allodynia were assessed for ipsilateral hindpaws in rats that received intraspinal injections of AAV5-GFAP-Fcgr2a-GFP (GFP) or AAV5-GFAP-Fcgr2a-Cas9 (Cas9) 18 days prior to CCI (800 nL, 8 × 10⁹ vg). Paw withdrawal thresholds were measured at baseline (BL) and at repeated intervals after CCI. (B) von Frey thresholds were measured for contralateral hindpaws within the same group of rats as in (A). n = 7–9 rats per group (2–6 males and 3–5 females). Two-way repeated measures ANOVA and Šidák’s post hoc test; *P < 0.05, **P < 0.01. (C) Representative 100x image of ipsilateral L4–5 spinal cord dorsal horn after Cas9 AAV intraspinal injection, collected after completion of behavioral testing 56 days after CCI. Immunofluorescence depicts viral transduction (visualized by HA-tag, in yellow) and GFAP⁺ astrocytes (pink). Scale bars indicate 100 μm. (D) Colocalization of HA-tag with GFAP was quantified with Manders’ coefficient (tM) in ipsilateral dorsal horn. n = 6 rats (2 males and 4 females). Data are presented as mean ± SEM. For individual data points, male subjects depicted as circles and female subjects depicted as triangles.

IgG immune complexes induce proinflammatory signaling in astrocytes via Fcgr2a

When IgG-IC binds to FcγRIIa, receptor crosslinking leads to phosphorylation of a cytosolic immunoreceptor tyrosine-based activation motif (ITAM). This can cause recruitment of spleen tyrosine kinase (Syk), initiating various downstream signaling cascades including translocation of nuclear factor kappa B (NF-κB) to drive transcription of inflammatory mediators^13,31. Based on our observations that FcγRIIa is expressed on astrocytes and contributes to the emergence of allodynia, we reasoned that engagement of glial FcγRIIa could lead to Syk-mediated transcription of proinflammatory cytokines and chemokines, whose release has been strongly linked to enhanced neuronal excitability and nociceptive hypersensitivity^32,33. In order to examine the outcomes of activating FcγRIIa exclusively on astrocytes, we first purified and cultured primary astrocytes through immunopanning to obtain cells with a transcriptional profile and morphology closely resembling in vivo astrocytes¹⁶. Immunocytochemistry confirmed that cultured cells were positive for GFAP and retained the complex branching morphology of mature astrocytes (Figure S3A–B), and RT-PCR was used to verify their enrichment of astrocyte genes (Figure S3C). To determine if FcγRIIa activation influences the inflammatory response of astrocytes, cells were first transfected with siRNA to knockdown expression of Fcgr2a, and subsequently treated with IgG-IC (3–10 μg/mL) to activate FcγRs. In astrocytes transfected with a negative control siRNA, IgG-IC elicited a dose-dependent increase of transcription of several canonical proinflammatory cytokines and chemokines, including Il1b (Figure 4A), Tnf (Figure 4B), Ccl2 (Figure 4C), and Cxcl1 (Figure 4D). The highest concentration of IgG-IC (10 μg/mL) induced significantly greater expression of these proinflammatory factors compared to media-treated astrocytes (Figure 4A–D). Knockdown of Fcgr2a by siRNA significantly reduced transcription of cytokines Il1b and Tnf in response to the highest concentration of IgG-IC (Figure 4A–B), and attenuated production of chemokines Ccl2 and Cxcl1 (Figure 4C–D). Thus, IgG-IC stimulates an FcγRIIa-dependent proinflammatory cytokine response in astrocytes.

Figure 4.

Activating astrocyte FcγRs with IgG-IC increases transcription of proinflammatory cytokines and chemokines, which is partially driven by Fcgr2a. (A to D) On DIV5, astrocytes were transfected with Fcgr2a siRNA or negative control siRNA. On DIV7, FcγRs on astrocytes were activated with IgG-IC (3–10 μg/mL), with 0 μg/mL representing media-only reference control. Relative gene expression for (A) Il1b, (B) Tnf, (C) Ccl2, and (D) Cxcl1 were quantified with RT-qPCR after 4-hour incubation. n = 3 wells per group. Kruskal-Wallis and Dunn’s post hoc test; *P < 0.05, **P < 0.01 compared to media-only control. (E to H) Astrocytes were pretreated with Syk inhibitor R406 (3 μM) or NF-κB inhibitor PDTC (30 μM) for 1 hour, and FcγRs were activated with IgG-IC (10 μg/mL) for 4 hours. Relative gene expression for (E) Il1b, (F) Tnf, (G) Ccl2, and (H) Cxcl1 were quantified with RT-qPCR. n = 3–6 wells per group. Kruskal-Wallis and Dunn’s post hoc test; *P < 0.05, **P < 0.01 compared to media-only control. (I to J) Astrocytes were pretreated with Syk inhibitor R406 (3 μM) or NF-κB inhibitor PDTC (30 μM) for 1 hour, and FcγRs were activated with IgG-IC (10 μg/mL) for 24 hours. Protein expression of IL-1β (I) and TNF (J) in cell culture supernatant were quantified by ELISA. n = 3–6 wells per group. Kruskal-Wallis and Dunn’s post hoc test; *P < 0.05, ***P < 0.001. Data are presented as mean ± SEM.

To clarify if IgG-IC induces proinflammatory signaling via Syk and NF-κB signaling downstream of FcγRIIa activation, astrocytes were pretreated with Syk inhibitor (R406) or NF-κB inhibitor (PDTC) before stimulating with IgG-IC (10 μg/mL). Both R406 and PDTC significantly reduced transcription of Il1b (Figure 4E), Tnf (Figure 4F), Ccl2 (Figure 4G), and Cxcl1 (Figure 4H) in response to IgG-IC. To quantify cytokine secretion after IgG-IC, we collected astrocyte culture supernatant 24 hours later and quantified protein using ELISA. We confirmed that IL-1β (Figure 4I) and TNF (Figure 4J) protein expression increased in response to IgG-IC treatment, and pretreatment with R406 or PDTC prevented this increase in cytokine release (Figure 4I–J). Taken together, these data suggest that FcγRIIa activation on astrocytes promotes Syk and NF-κB-mediated intracellular signaling that culminates in increased production of proinflammatory cytokines and chemokines.

Discussion

We demonstrate that the IgG antibody receptor FcγRIIa on spinal cord astrocytes contributes to the development of mechanical allodynia following peripheral nerve injury in male and female rats. Transcription of Fcgr2a was dynamically and discretely upregulated in ipsilateral spinal cord dorsal horn after CCI, while no changes in Fcgr2a expression were observed in the lumbar DRGs, contralateral lumbar spinal cord, or the ipsilateral cervical spinal cord. These results imply that injury-induced changes in Fcgr2a are localized to cells within the ipsilateral spinal cord. Immunofluorescence revealed FcγRIIa expression increased in the spinal cord following CCI and was primarily found on GFAP⁺ astrocytes. Based on these observations, we hypothesized that astrocytic FcγRIIa may contribute to nerve injury-induced pain via secretion of inflammatory mediators. Using custom viral CRISPR-Cas9 gene editing constructs, we demonstrated that knockdown of Fcgr2a attenuated mechanical allodynia after injury. Using primary cultures of astrocytes, we established that activation with IgG-IC caused production of pronociceptive inflammatory cytokines and chemokines that was partially dependent upon Fcgr2a expression. Based on these results, we propose that neuropathic injury results in the production of autoreactive IgG which activates FcγRIIa in astrocytes, promoting production of proinflammatory cytokines like IL-1β and TNF. These cytokines are critical for the maintenance of neuron hyperexcitability and mechanical allodynia^32–35. Collectively, our results implicate astrocytic FcγRIIa signaling as a previously unappreciated contributor to spinal neuroinflammation in the manifestation of allodynia following neuropathic injury.

Nerve injury led to a time-dependent and spatially restricted increase in Fcgr2a expression in the ipsilateral L4–5 dorsal spinal cord. These results align with previous evidence that nerve ligation models are associated with increased expression of FcγRs in the CNS^9,14. Indeed, subsequent protein-protein interaction network analysis of a spinal cord microarray dataset¹⁴ revealed a prominent interaction network for genes associated with antibody receptor signaling, including Fcgr1a, Fcgr2a, Fcgr2b, and Fcer1g encoding FcRγ, a key subunit for FcγRs⁶. Spinal cord injury similarly increases Fcgr2a, Fcer1g, and Syk transcription in rat spinal cords months after contusion, indicating both peripheral and central lesions can influence Fcgr2a expression³⁶. We did not identify any changes in Fcgr2a after CCI in contralateral L4–5 or ipsilateral C3–4 dorsal spinal cord, nor did we detect changes in the ipsilateral or contralateral L4–5 DRGs. This suggests that Fcgr2a transcription may be altered only within tissue receiving input from the injured sciatic nerve, raising the question of which cues are responsible for this increase. One possibility is that input-specific alterations in neuronal activity from injured and hyperactive nociceptors could trigger increased Fcgr2a expression at cells within the ipsilateral spinal cord. Indeed, we observed increased immunoreactivity for FcγRIIa primarily in the superficial laminae of the spinal dorsal horn, where the majority of nociceptors (both unmyelinated C-fibers and thinly myelinated Aδ-fibers) terminate³⁷. We speculate that this regionally selective increase in Fcgr2a may explain how autoreactive IgG can promote nociceptive hypersensitivity specifically in the ipsilateral hindpaw, despite elevated autoreactive IgG in the circulation⁸.

Microglia and neurons have been shown to increase expression of FcγRs after peripheral nerve injury, suggesting multiple cell types may modulate antibody receptor expression in neuropathic pain^9,38,39. Previous research established that Fcgr2a expression in the dorsal spinal cord correlates with allodynia severity following CCI, although its relative expression among cell types was previously unknown¹⁴. Here, we found FcγRIIa was predominantly expressed by GFAP⁺ astrocytes. Relatively little is known about the function of astrocytic FcγRs. Astrocytes cultured from the developing rat brain express transcripts for all FcγRs, including Fcgr2a, but protein levels had not previously been assessed⁴⁰. Given that elevated GFAP localized specifically in the ipsilateral dorsal horn is a hallmark of neuropathic injury³⁵, we surmise that IgG signaling at FcγRIIa could account for sustained and spatially restricted astrocyte activation after nerve injury.

Neuroinflammation from astrocytes and microglia have long been associated with the initiation and maintenance of neuropathic pain^35,41, and astrocytes are capable of producing cytokines and chemokines to induce or prolong inflammation and gliosis associated with nerve injury-induced mechanical allodynia and chemotherapy-induced peripheral neuropathy^42–44. Spinal astrocytes release of IL-1β and TNF induces long-term potentiation of synapses between nociceptors and neurons in the dorsal horn, demonstrating that astrocytes can influence the activity and plasticity of neural circuits encoding pain³³. Cytokines released by astrocytes can act directly on cytokine receptors expressed by neurons to modulate their activity, and there is evidence for neuronal expression of IL-1 receptor^45–47, TNF receptor⁴⁸, CCL2 receptor (CCR2)⁴⁹, and CXCL1 receptor (CXCR2)^50,51. Optogenetic activation of channelrhodopsin-2 (ChR2)-expressing spinal astrocytes enhances cytokine transcription of Il1b, Tnf, Ccl2, and Il6 in the ipsilateral spinal cord with concomitant elicitation of mechanical hypersensitivity, demonstrating that astrocytes can directly influence nociception via release of these inflammatory mediators⁵². Here, we present evidence that IgG-IC elicits transcription of Il1b, Tnf, Ccl2, and Cxcl1 in an FcγRIIa-dependent manner. FcγR crosslinking by IgG-IC typically leads to phosphorylation of intracytoplasmic ITAMs by Src family kinases and recruitment of Syk kinases, which in turn activates downstream signaling pathways involved in cytokine production^12,13. Indeed, we demonstrated that IgG-IC triggers proinflammatory transcription in astrocytes through Syk and downstream NF-κB activation. This supports previous observations that neuronal FcγR activation by IgG-IC elicits a Syk-mediated signaling cascade to enhance neuronal depolarization⁵³. We propose that FcγR-mediated release of proinflammatory cytokines and chemokines from spinal astrocytes could promote allodynia, although additional in vivo experiments are needed to support this hypothesis.

Considering FcγRIIa is activated by immune complexes (formed by autoantigens and autoreactive IgG released by B cells), our results are in line with previous research linking B cells, autoreactive IgG, and FcγR signaling to pain^6,30,54–59. Chronic pain often accompanies autoimmune diseases characterized by circulating autoantibodies, including rheumatoid arthritis, complex regional pain syndrome (CRPS), Guillain-Barré syndrome, and systemic lupus erythematosus⁶⁰. Autoantibodies play a causal role in pain associated with neurological disorders^61,62 and paraneoplastic neuropathy^63,64, and autoreactive antibodies have been well-characterized as contributing to pain in preclinical models and passive transfer experiments with rheumatoid arthritis^17,55,56, CRPS^65–68, fibromyalgia^69,70 and intervertebral disc injury^71,72. IgG accumulates in the ipsilateral lumbar DRGs and spinal cord of mice after CCI, and a similar pattern of IgG deposition was observed in DRGs of humans with a history of chronic pain⁶. Passive transfer experiments reveal that IgG from CCI mice induces mechanical allodynia in B cell-deficient mice and depends upon intact FcγR signaling⁶. Beyond neuropathic pain, B cells activation occurs in a mouse model of chronic musculoskeletal pain⁷³, and evidence of B cell enrichment has been observed in the cervical DRG of a patient with spreading neck pain⁷⁴ and the entrapped lesioned nerves from patients with Morton’s neuroma⁷⁵. Therefore, our current results support the premise that B cells and their production of autoantibodies are an important conduit for promoting pain in conditions not typically defined by autoimmunity, such as neuropathic pain⁷⁶.

There are several limitations of this work that should be acknowledged. After intraspinal viral injections, the expression of the HA-tag extended beyond astrocytes. Although we showed that FcγRIIa was most prominently expressed by spinal astrocytes, we cannot exclude the possibility that knockdown in other cell types may contribute to attenuation of allodynia. However, our in vitro studies support the notion that astrocytic FcγRIIa signaling contributes to allodynia, where stimulation of this receptor resulted in production and release of pro-nociceptive factors. A limitation of the in vitro studies is that astrocytes were cultured from postnatal brain according to validated protocols²⁸, which may not be entirely reflective of spinal astrocytes. Immunopanning achieves high purity cultures of postmitotic astrocytes whose transcriptional profile and branching morphology more closely resembles mature astrocytes¹⁶. Despite these advantages, this approach is not optimized for purifying spinal astrocytes due to high oligodendrocyte contamination and low cell yield. Further work is needed to overcome this limitation, which could generate important insight into regional-specific features of spinal astrocytes.

As researchers have sought to uncover the mechanisms governing the emergence of neuropathic pain, substantial preclinical and clinical observations have revealed how neuroimmune interactions shape the trajectory of whether pain persists or resolves^34,77,78. In this study, we demonstrate that FcγRIIa on spinal astrocytes contribute to the development of nerve injury-induced tactile pain. Conditional knockdown of Fcgr2a in GFAP-expressing astrocytes prevented the development of mechanical allodynia, which may result from inflammatory cytokine release in the ipsilateral dorsal spinal cord. Our results elaborate upon a growing set of evidence that FcγRs, traditionally associated with the adaptive immune system and autoimmunity, can also influence nociception based on its expression in the CNS. Modulating FcγR signaling, as recently shown using pan IgG Fc receptor interacting molecule (PRIM)⁷¹, warrants further investigation as a therapeutic strategy to alleviate neuropathic pain.

Highlights.

• Spinal astrocytes upregulate FcγRIIa expression following peripheral nerve injury in rats

• CRISPR-Cas9–mediated disruption of Fcgr2a in astrocytes reduces mechanical allodynia.

• IgG immune complexes trigger astrocytic FcγRIIa–Syk–NFκB signaling for proinflammatory responses.

• Astrocyte FcγRIIa signaling is a previously unappreciated mechanistic contributor to neuropathic pain

Perspective.

Activation of FcγRIIa signaling in spinal astrocytes promotes mechanical allodynia following nerve injury and initiates neuroinflammatory pathways in response to IgG immune complexes. These findings reveal autoantibody IgG signaling at glial Fcγ receptors as a potential therapeutic approach to alleviate neuropathic pain.

Data availability

Data are available from the corresponding author upon request.