Astrocytes coordinate neuronal signaling in physiological conditions but can also drive neuroinflammation in pathophysiologic conditions, such as chronic pain. How and when astrocyte molecular pathways change in response to pain-inducing peripheral injury is key to understanding the acute-to-chronic pain transition. Here, we utilize translating ribosome affinity purification technology in a mouse model of complex regional pain syndrome to uncover the functional molecular signature of spinal astrocytes early and late post-injury. We find that astrocytes exhibit a temporally distinct translatome with most significant gene expression changes occurring acutely after injury. We further identify astrocyte lipid metabolism as altered after injury and demonstrate that lipid droplets (marked by PLIN2) accumulate in the spinal dorsal horn in the chronic post-injury phase. Overall, this work provides an astrocyte-specific translatome resource for understanding spinal astrocyte contributions to pain and highlights spinal cord lipid metabolism as a pathway of interest in pain pathophysiology.

Mice

Adult male and female mice, 10-14 weeks old at the start of the experiment,s were housed 2-5 per cage and maintained on a 12-hour light/dark cycle in a temperature-controlled environment with ad libitum access to food and water. Male mice weighed approximately 25 g at the start of the study, and female mice weighed approximately 20 g at the start of the study. Mice used in this study: wild type C57BL/6J (Jax stock #00664), Aldh1l1-CreERT2 (Jax stock #029655), and Rpl22-HA (Jax stock #029977). To specifically and conditionally express the hemagglutinin epitope on the Rpl22 ribosomal subunit in astrocytes, we crossed hemizygous Aldh1l1-CreERT2 with homozygous Rpl22-HA and then subsequently back-crossed to homozygous Rpl22-HA. Thus, experimental animals were homozygous for the Rpl22-HA allele and hemizygous for the CreERT2 knock-in allele. Genotypes were determined using RT-PCR with probes designed for each gene (Transnetyx, Cordova, TN).

Study Approval

All procedures were approved by the Stanford University Administrative Panel on Laboratory Animal Care and the Stanford University Institutional Animal Care and Use Committee in accordance with American Veterinary Medical Association guidelines and the International Association for the Study of Pain.

Drugs and Route of Administration

Tamoxifen (Sigma, #T5648) was dissolved in corn oil at a concentration of 25 mg/mL, and mice were injected intraperitoneally (i.p.) at 100 mg/kg daily for 5 consecutive days. Spinal cord was collected for immunoprecipitation and immunohistochemistry experiments 5-6 days after the last dose of tamoxifen to allow for sufficient recombinase activity.

Tibial Fracture-Cast model of Complex Regional Pain Syndrome (CRPS)

Surgeries were performed as previously described. Briefly, mice were anesthetized with isoflurane and a hemostat was used to make a closed fracture of the right distal tibia. The hind limb was then wrapped in casting tape (ScotchCastTM Plus) from the metatarsals of the hind paw up to a spica formed around the abdomen. The cast around the paw included a window left open over the dorsum of the paw and ankle to prevent constriction when post-fracture edema developed. Mice were inspected throughout the post-operative period of cast immobilization to ensure proper cast position. At 3 weeks post-fracture, mice were briefly anesthetized with isoflurane and casts were removed with cast shears. Mice were given 48-72 hours of recovery time before the start of behavioral assessment. 

Behavioral testing

All in vivo behavioral testing was conducted between 8:00 am and 2:00 pm in an isolated, temperature- and light-controlled room. Mice were acclimated for 30-60 minutes in the testing environment in custom clear plastic cylinders (4” D) on a raised metal mesh platform (24” H), with an additional 20-30 minutes of acclimation with the blinded experimenter in the room. To ensure blinding to experimental groups, mice were randomly placed in a cylinder by a second experimenter, and mouse identification numbers were recorded after testing was complete.

Paw edema

Hind paw edema was determined by measuring the hind paw dorsal-ventral thickness over the midpoint of the third metatarsal with a digital caliper (Avantor, #62379-531) while the mouse was briefly anesthetized with isoflurane. Data were analyzed as the difference between the injured and uninjured side and averaged across experimental groups. Hind paw edema was measured at 3- and 5-week post-fracture.

Unweighting

An incapacitance device (IITC Inc Life Science, #600M) was used to measure hind paw unweighting. Mice were positioned vertically in a plastic mouse holder (IITC Inc Life Science, #600MH) facing away from the experimenter and with the hind paws resting on separate metal scale plates. The entire weight of the mouse was supported on the hind paws, and the tail was adjusted off the scale plates as needed by the experimenter. The duration of each measurement was 4-6 seconds, and 6 consecutive measurements were taken at 60-second intervals. Six readings were averaged to calculate the bilateral hind paw weight-bearing values, and data were analyzed as a ratio between the right hind paw weight divided by the sum of right and left hind paw values [2R/(R+L)] x 100. Unweighting was measured at 3- and 5-week post-fracture.

Mechanical nociception assay

To evaluate mechanical reflexive hypersensitivity, we used a logarithmically increasing set of 9 von Frey filaments (Stoelting, #58011), ranging in force from 0.02 to 4.0 g. These were applied perpendicular to the plantar hind paw with sufficient force to cause a slight bending of the filament. A positive response was characterized as a rapid withdrawal of the paw away from the stimulus filament. Using the up-down statistical method, the 50% withdrawal mechanical threshold scores were calculated for each mouse and averaged across the experimental groups. Mechanical nociception testing was performed at baseline (pre-fracture) and at weeks 3, 5, and 7 post-fracture.

Ribosome Immunoprecipitation and RNA Isolation

Mice were deeply anesthetized with isoflurane at 0 (uninjured), 3-, 5-, and 7-weeks post-fracture. Spinal cord tissue was isolated through hydraulic extrusion and the lumbar region ipsilateral to injury (the right half of the lumbar region when the dorsal portion of the cord is facing up) was quickly dissected by halving the cord down the midline with a scalpel on a pre-chilled RNase-free metal block on ice and under a dissection microscope. 2-4 ipsilateral lumbar cords were pooled per sample (Supp. Figure 1). The immunoprecipitation protocol was followed as described in Sanz et al., 2009. Briefly, pooled cords were dounce-homogenized at <3% weight/volume in 1mL of ice-cold supplemented homogenization buffer (50 mM Tris, pH 7.5, 100 mM KCl, 12 mM MgCl2, 1% NP-40, 1 mg/mL heparin, 100 μg/mL Cycloheximide, 1mM DTT, 200 U/mL RNAsin Plus, and 1X Protease inhibitor). Homogenates were then centrifuged at 10,000 rpm for 10 minutes at 4°C. The supernatant was collected and incubated with anti-HA antibody (BioLegend, #901502, 1:100) for 4 hours at 4°C with gentle rotation. Prior to antibody incubation, 100 uL of the supernatant was mixed with 300 uL of TRIzol LS (Invitrogen, #10296010) and stored at -80°C for future use as the respective “Input” sample, which represents total cellular RNA from non-immunoprecipitated homogenate. Following antibody incubation, homogenates were incubated with Dynabeads Protein G (Invitrogen, #10004D) at 4°C overnight with gentle rotation. The following day, homogenates were placed on a magnetic rack (Invitrogen, #12321D) at 4°C for 1 minute, supernatant was removed, and the remaining bead-antibody-ribosome complexes were washed with high-salt buffer (50 mM Tris, pH 7.5, 300 mM KCl, 12 mM MgCl2, 1% NP-40, 1 mM DTT, 100ug/mL Cycloheximide, dH₂O to achieve final volume). This wash step was repeated a total of 5 times. After the last wash, bead-antibody-ribosome complexes were mixed with TRIzol (Invitrogen, #15596018), vortexed rigorously for 30 seconds to release the ribosome-bead bond, and stored at -80°C for RNA extraction. These final samples were used as the immunoprecipitated or “IP” samples. RNA was isolated according to the manufacturer’s protocol using the Zymo RNA Clean & Concentrator kit (Zymo Research, #R1015), and RNA concentration and purity were determined using either NanoDrop 2000 or Agilent 2100 Bioanalyzer.

Sample Preparation for Sequencing

A total of 32 samples, equally split between male and female mice, were submitted to Azenta Life Sciences (formerly Genewiz) for RNA isolation and bulk RNA sequencing. Of these, 24 were immunoprecipitated (IP) sampl,es and 8 were representative Input samples. All samples were sequenced on an Illumina NovaSeq using an unstranded protocol to generate 150bp paired-end reads. Five samples of low RNA quality (RIN < 7) were removed from further analyses.

RNASeq Data Processing and Analysis

FASTQ files were passed through the nf-core RNAseq (v3.12.0) pipeline, which performed quality control (FastQC), trimming (Trim Galore), alignment to genome reference (STAR; GRCm38), and transcript quantification (RSEM). Genes with ≥ 25 counts in at least 90% of samples were included for analysis. Sample gene count matrices were then transformed using a variance-stabilizing transformation and batch-corrected. Sex and batch were collinear, and as such while sex was not directly adjusted for, sex-associated signals are likely to be lost with batch correction. Processed samples were passed through weighted gene co-expression network analysis (WGCNA; v2.1.0). WGCNA analysis was performed using a module discrimination threshold of 0.2 (MEDissThres). Recovered modules were correlated to timepoint conditions with the baseline samples as the comparison cohort. Correlation p-values were FDR corrected, and only significantly associated modules were further investigated (Pearson R >= 0.7; p~-adjusted ~<= 0.05). No modules were significantly correlated with batch or sex. Module membership and gene significance were calculated for each gene. As a supplementary analysis, gene significance was also calculated based on the raw transcript counts and compared against the gene significance values calculated from the processed counts. For all modules, the Ordinary least squares regression R² was > 0.98 and the slope > 1, suggesting that variance-stabilizing transformation and batch correction strengthened the signal-to-noise ratio of the dataset (Supp. Figure 4). Lastly, using an external mouse spinal cord single-cell RNA-seq dataset, we extracted astrocyte gene expression specificity by computing the log2 fold change in expression between labeled astrocyte populations vs. the rest of the populations using Scanpy(v1.10.4).

Gene Ontology and Signaling Pathway Analysis

Functional enrichment analysis was performed using g: Profiler with Benjamini-Hochberg multiple testing correction with a significance threshold of FDR>0.05. Functional analyses were generated through the use of Ingenuity Pathway Analysis (QIAGEN Inc., https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis).

Immunohistochemistry

Mice (10-20 weeks) were transcardially perfused with PBS and 10% formalin in PBS (except for PLIN2 immunofluorescence; tissue collection and processing are described below). Spinal cords were harvested by hydraulic extrusion and post-fixed in 10% formalin for 4 hours at 4°C. Spinal cords were then cryoprotected in 30% sucrose in PBS and frozen in O.C.T. (Tissue-Tek, #4583), and 40 μm sections were prepared using a cryostat (Leica Biosystems). Free-floating spinal cord sections were incubated in blocking solution (5% normal donkey serum and 0.3% Triton X-100 in PBS) for 1 hour at room temperature, followed by incubation with primary antibodies at 4°C, overnight. The following primary antibodies were used: rabbit anti-HA (Cell Signaling, #3724, 1:500), and mouse anti-GFAP (Sigma, #G3893, 1:400). Sections were washed with 1X PBS 3 times and incubated with the appropriate secondary antibody conjugated to AlexaFluor (1:1000) in 1% normal donkey serum, 0.3% Triton X-100, and 1X PBS for 2 hours at room temperature. Sections were then mounted onto slides (ThermoFisher, #12-550-15) with Fluoromount G with DAPI medium (ThermoFisher, #00-4959-52).

For PLIN2 immunofluorescence, mice (10-20 weeks) were transcardially perfused with PBS and 4% PFA in PBS. Spinal cords were similarly harvested by hydraulic extrusion and post-fixed in 4% PFA for 8 hours,s shaking at room temperature. Spinal cords were then cryoprotected, sectioned, and stained as described above using rabbit anti-Plin2 (Proteintech, #15294-1-AP, 1:500).

Image Acquisition and Quantification of HA Colocalization

Z-stacks were taken using a 40X (NA 0.95) objective on a Keyence BZ-X800 fluorescent microscope. Each z-stack was taken at the same exposure using 30 slices with a 0.4 um step size, and five lumbar spinal cord dorsal horn sections were imaged and averaged per mouse. ImageJ/FIJI software was used to quantify colocalization between HA, GFAP, and Iba1 staining. To determine the percent specificity of HA expression in astrocytes, the number of HA/GFAP double-positive cells was divided by either the total number of HA+ or GFAP+ cells. All imaging and quantification were performed in a blinded manner.

Image Acquisition and Lipid Droplet Quantification

Z-stacks were taken using a 20x (NA 0.75) objective on a Keyence BZ-X800 fluorescent microscope. Each z-stack was taken at the same exposure using 30 slices with a 0.5 um step size, and 3-5 lumbar spinal cord dorsal horn sections were imaged and averaged per mouse. ImageJ/FIJI software was used to analyze lipid droplet particles. For each section, the gray matter region of the spinal cord dorsal horn was manually delineated using the freehand ROI tool, with anatomical boundaries defined consistently across samples. A threshold was applied uniformly across images (min. value = 40) to isolate lipid droplet signal while minimizing background fluorescence. Lipid droplets (LDs) within the defined dorsal horn ROI were quantified using the Analyze Particles function. Particle size was restricted to 0.2–3.5 μm² to exclude background noise and non-droplet structures. For each ROI, this analysis yielded the total LD count and the percentage of the ROI area occupied by LDs. LD density was calculated by normalizing LD counts to the ROI area per slice of spinal cord (number of LDs per μm²). All analysis parameters were held constant across experimental conditions to allow direct comparison between samples.

QUANTIFICATION AND STATISTICAL ANALYSIS

Cohort sizes were determined based on historical data from our laboratory using a power analysis to provide >80% power to discover 25% differences with p<0.05 between groups, requiring a minimum of 4 animals per group for all behavioral outcomes. All experiments were randomized by cage and performed by a blinded researcher. Researchers remained blinded throughout histological and behavioral assessments and were unblinded at the end of each experiment before statistical analysis. All data are presented as mean ± SEM unless indicated. Statistical analysis was performed using GraphPad Prism version 10.4.1 (GraphPad Software) or Python. Data were analyzed using an unpaired t-test, mixed effects analysis with Sidak’s post hoc test, ordinary one-way with Dunnett’s post-hoc test or Tukey’s post-hoc test, or two-way analysis of variance with Tukey’s post-hoc test, as indicated in the main text or figure captions. The ‘‘n’’ for each experiment is listed in the figure legends. No data were excluded from histological or behavioral analyses.