Formoterol alters chemokine expression and ameliorates pain behaviors after moderate spinal cord injury in female mice

Abstract

Secondary spinal cord injury (SCI) is characterized by increased cytokines and chemokines at the site of injury that have been associated with the development of neuropathic pain. Nearly 80% of SCI patients report suffering from chronic pain, which is poorly managed with available analgesics. While treatment with the US Food and Drug Administration-approved β₂-adrenergic receptor agonist formoterol improves various aspects of recovery post-SCI in vivo, its effects on cytokines, chemokines, and neuropathic pain remain unknown. Female mice were subjected to moderate (60 kilodynes [kdyn]) or severe (80 kdyn) SCI followed by daily treatment with vehicle or formoterol (0.3 mg/kg, i.p.) beginning 8 hours after injury. The expression of proinflammatory cytokines/chemokines, such as interferon gamma-induced protein 10, macrophage inflammatory protein 1a, monocyte chemoattractant protein 1, B-cell attracting chemokine 1, and nuclear factor kappa-light-chain-enhancer of activated B-cells, was increased in the injury site of vehicle-treated mice 24 hours post-SCI, which was ameliorated with formoterol treatment, regardless of injury severity. Thermal hyperalgesia and mechanical allodynia, as measured by Hargreaves infrared apparatus and von Frey filaments, respectively, were assessed prior to SCI and then weekly beginning 21 days post-injury (DPI). While all injured mice exhibited decreased withdrawal latency following thermal stimulation compared with baseline, formoterol treatment reduced this response ∼15% by 35 DPI. Vehicle-treated mice displayed significant mechanical allodynia, as evidenced by a 55% decrease in withdrawal threshold from baseline. In contrast, mice treated with formoterol maintained a consistent withdrawal time at all times tested. These data indicate that formoterol reduces inflammation post-SCI, likely contributing to mitigation of neuropathic pain and further supporting the therapeutic potential of this treatment strategy.

Significance Statement

Chronic pain is a detrimental consequence of spinal cord injury (SCI). We show that treatment with the US Food and Drug Administration-approved drug formoterol after SCI decreases injury site proinflammatory chemo-/cytokines and alters markers of glial cell activation and infiltration. Additionally, formoterol treatment improves locomotor function and body composition, and decreases lesion volume. Finally, formoterol treatment decreased mechanical allodynia and thermal hyperalgesia post-SCI. These data are suggestive of the mechanism of formoterol-induced recovery, and further indicate its potential as a therapeutic strategy for SCI.

1. Introduction

Spinal cord injury (SCI) is a debilitating condition with an annual occurrence of nearly 18,000 new cases in the United States each year (National Spinal Cord Injury Statistical Center, 2022). Up to 80% of SCI sufferers experience chronic pain after injury (Yezierski, 2000). Neuropathic pain, the most prevalent category of post-SCI pain (Kerr and David, 2007), is caused by abnormal communication between damaged nerves and the brain resulting in amplified responses to innocuous and/or noxious stimuli (Jensen et al, 2001; Costigan et al, 2009). This is often reported as being the most prominent factor for decreased patient quality of life post-SCI. In fact, approximately 40% of SCI patients suffering from pain would be willing to exchange the chance of any functional recovery for pain relief (Yezierski, 2000; Mehta et al, 2013). While there are approved therapeutic options to reduce pain, neuropathic pain is particularly resistant to treatment with current analgesics (Miranpuri et al, 2021). Furthermore, these analgesics do not improve other aspects of recovery post-SCI, speaking to the necessity for continued research into alternative and comprehensive therapeutic strategies.

The secondary injury cascade following SCI is characterized by vascular disruption, which decreases oxygen delivery at the injury site, directly impacting mitochondrial function and homeostasis (Scholpa and Schnellmann, 2017). While the mechanism is not fully understood, some studies implicate mitochondrial dysfunction in the pathogenesis of neuropathic pain (Joseph and Levine, 2006; Sui et al, 2013; Dai et al, 2020). Recently, reactive oxygen species (ROS) have been implicated in chronic pain, and oxidative stress in the spinal cord may play a role in hyperalgesia (Schwartz et al, 2008; Schwartz et al, 2009; Dai et al, 2020). Various additional mitochondrial functions, including opening of the mitochondrial permeability transition pore, have also been connected to neuropathic pain (Dai et al, 2020). Mitochondrial biogenesis (MB) is a complex transcriptional program that increases functional mitochondria (Ventura-Clapier et al, 2008). Given that studies have demonstrated a role for spinal cord mitochondrial function in models of persistent hyperalgesia (Schwartz et al, 2009), therapeutics targeting reestablishment of mitochondrial function through increased MB could stimulate multiple aspects of recovery post-SCI (Scholpa and Schnellmann, 2017), including chronic pain.

Neuroinflammation is a well characterized hallmark of secondary injury and involves the activation of glial cells, which leads to the production of cytokines and chemokines that further promote the inflammatory cascade. Continued activation and secretion of proinflammatory molecules has been strongly implicated in neuronal sensitization and the formation of neuropathic pain after SCI (DeLeo and Yezierski, 2001; Chen et al, 2012). In fact, levels of neuroinflammatory metabolites have been shown to be elevated in SCI patients with neuropathic pain, with higher magnitude of neuroinflammation being correlated with greater pain sensation (Pfyffer et al, 2020).

Other than surgical approaches, the treatment options available for pain post-SCI are palliative, such as anticonvulsants, opioids, and anesthetics (Kanpolat et al, 2008; Mehta et al, 2013). Despite the availability of these medications, many have broad modes of actions and unfavorable side-effects. Furthermore, studies indicate that these interventions provide only 50% pain reduction to only one-third of sufferers post-SCI (Siddall, 2009). We have previously shown that treatment with the mitochondrially biogenic US Food and Drug Administration (FDA)–approved β₂-adrenergic receptor (ADRB2) agonist formoterol after severe contusion SCI increases functional and mitochondrial recovery (Scholpa et al, 2019a,b, 2021). Importantly, formoterol treatment improved neuropathic pain in models of paclitaxel-induced pain (Chen et al, 2021), spared nerve injury (Damo et al, 2023), and sciatic nerve cuffing (Ceredig et al, 2019; Kremer et al, 2020). The effect of formoterol on neuropathic pain post-SCI, however, remains unknown. To address this gap in knowledge, this study examined the effect of formoterol on neuroinflammation and neuropathic pain development following SCI using thermal hyperalgesia and mechanical allodynia.

2. Materials and methods

2.1. Animals and SCI model

SCI was performed as described previously (Scholpa et al, 2019a,b, 2021; Scholpa, 2023). Briefly, female wild-type C57Bl/6 J mice 7–8 weeks of age from The Jackson Laboratories were housed in groups of 3–5, and allowed to acclimate for 7 days prior to use. Mice were randomized into sham and SCI groups and anesthetized with ketamine/xylazine (10 mg/kg/6 mg/kg) via intraperitoneal (i.p.) injection, followed by complete single-level laminectomy at the 11th thoracic vertebrae. The vertebral column was clamped and stabilized at the upper thoracic and lumbar levels and a controlled contusion with a force of either 60 or 80 kilodynes (kdyn) and 0 seconds dwell time was administered using the Infinite Horizon IH-0400 impactor with the dura intact (Scholpa et al, 2019b). Sham mice received the laminectomy only. Injured mice were divided into vehicle- or formoterol-treated groups and manual bladder expression was performed 2 times per day. Impact statistics can be found in Table 1. All animals were weighed prior to surgery, 24 hours post-SCI, and every other day thereafter until euthanasia via anesthesia overdose. Spinal cords were isolated, and gastrocnemius muscles were removed and weighed. All studies were approved by the Institutional Animal Care and Use Committee of the University of Arizona in accordance with the guidelines set forth by the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Table 1.

Impact statistics

	SCI + Vehicle			SCI + Formoterol
	Force (kdyn)	Displacement (μm)	Velocity (mm/s)	Force (kdyn)	Displacement (μm)	Velocity (mm/s)
60 kdyn	63.6	640.4	123.5	62.5	669.0	123.0
SEM	0.72	25.2	0.35	0.54	29.1	0.26
80 kdyn	83	970.6	122.3	84.2	914.4	122.4
SEM	1.16	26.3	0.17	0.86	18.1	0.14

2.2. Drug treatment

Formoterol fumarate dihydrate was obtained from Sigma-Aldrich (50-1742-902) and dissolved in <1% DMSO/saline to 0.03 mg/mL. Mice were treated i.p. daily with 0.3 mg/kg formoterol or vehicle beginning 8 hours post-SCI.

2.3. BioSpyder TempO-Seq and GO pathway analysis

Approximately 5 mm of spinal cord surrounding the site of injury was extracted 24 hours after SCI and RNA isolated via TRIzol per the manufacturer’s protocol (Invitrogen, 15596018,). Isolated RNA was purified via the Qiagen RNeasy MinElute Cleanup Kit (74204) and combined with lysis buffer provided by BioSpyder, as described previously (Scholpa, 2023). Samples were annealed to match pairs of detector oligos targeting a single gene with a detector oligo pool specific to the mouse transcriptome. Then, excess oligos were removed by nuclease digestion and the annealed oligos ligated to form amplifiable templates. Ligated oligos were amplified with primers with a different combination of tag sequences assigned to each sample and products were pooled into a sequencing library, which was purified and run on an Illumina Next-Gen Sequencer to count the number of sequences per sample. The resulting sequencing read counts from the library were then analyzed to determine the abundance of each gene in each sample (Yeakley et al, 2017; Scholpa, 2023). Gene set enrichment analysis was used to identify significantly altered Gene Ontology (GO) pathways and Cytoscape was used to identify altered networks and clusters via the AutoAnnotate application.

2.4. Milliplex cytokine/chemokine magnetic bead panel

Approximately 5 mm of spinal cord surrounding the site of injury was extracted 24 hours after SCI. Protein was isolated from the injury site using RIPA buffer (50 mM Tris-HCL, 150 mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton X-100, pH 7.4) with protease inhibitor cocktail (1:100), 1 mM sodium fluoride, and 1 mM sodium orthovanadate (Sigma-Aldrich) as described previously (Scholpa et al, 2019a). Samples were briefly sonicated and agitated on a rotator for 2 hours at 4 °C, then centrifuged at 14,000 × g for 15 minutes. The supernatant was then collected, and the pellet discarded. Protein was quantified using a bicinchoninic acid assay and diluted to 2 mg/mL. Total cytokine/chemokine concentrations were measured using a mouse multiplex immunoassay kit (MCYTOMAG-70K-PMX) with Luminex technology (MilliporeSigma), according to the manufacturer’s protocol and recommendations. Each sample, standard, and quality control were measured in duplicate. Plates were read using a MAGPIX instrument (Luminex) and results were analyzed using MILLIPLEX Analyst 5.1 software (MilliporeSigma).

2.5. Immunoblot

Approximately 5 mm of spinal cord surrounding the site of injury was extracted 24 hours after SCI and protein was isolated from the injury site using RIPA buffer plus inhibitors as described above. After quantification, 10 μg of protein was separated via electrophoresis using 4–15% SDS-PAGE gels and then transferred to nitrocellulose membranes (Bio-Rad). Membranes were cut surrounding expected molecular weights to probe for multiple targets. Given the small size of the tissue of interest, this allows for the preservation of sample and minimal, if any, stripping. Membranes were blocked with 5% BSA (Sigma-Aldrich, A4503-100G) or milk in TBST and incubated at 4 °C overnight with primary antibodies with constant agitation. Membranes were then washed and incubated with the appropriate horseradish peroxidase-conjugated secondary antibody followed by visualization using chemiluminescence (Thermo Scientific, 34075) on a GE ImageQuant LAS4000 (GE Life Sciences). Optimal density was analyzed using Image Studio Lite software. Primary antibodies used were as follows: monocyte chemoattractant protein 1 (MCP-1) (1:500, ThermoFisher PIPA5115555), nuclear factor kappa-light-chain-enhancer of activated B-cells (NF-ρB) (1:500, ThermoFisher, 51-0500), B-cell attracting chemokine 1 (BCA-1) (1:500, Abcam, ab190662), granulocyte colony-stimulating factor (G-CSF) (1:500, Abcam, ab181053), glutamate transporter 1 (GLT-1) (1:1,000, ThermoFisher, PA5-117492), transmembrane protein 119 (TMEM119) (1:1,000, Abcam, ab209064), glial fibrillary acidic protein (GFAP) (1:10,000, Abcam, ab4674), allograft inflammatory factor 1 (Iba1) (1:1,000, ThermoFisher, NC9288364), alpha tubulin (1:10,000, Abcam, ab52866). Density of target antibodies were normalized to alpha tubulin for each sample, followed by normalization to sham controls to obtain data as fold of sham (Supplemental Figs. 1 and 2).

2.6. Lesion volume analysis

Lesion volume analysis was performed as described previously (Scholpa et al, 2019a,b). Briefly, mice were transcardially perfused with 0.1 M phosphate buffered saline followed by 4% paraformaldehyde and a 6 mm segment centered on the injury site was frozen in optimal cutting temperature compound at −80 °C, followed by cryosectioning into 10 μm coronal sections. Eriochrome cyanine (Sigma-Aldrich, E2502) staining for myelin was used to distinguish damaged and spared tissue. Analyses were performed in a blinded fashion using a Motic EasyScan One slide scanner and Motic DSAssistant software (Schertz, TX). Lesion volume, percent spared tissue, and total volume were quantified across 2 mm of spinal cord centered on the epicenter at 100 μm intervals using the Cavalieri method (Scholpa et al, 2019a,b), totaling 21 sections per animal. Spared tissue was identified by both the cyto-architecture and positive staining of myelin in white matter, while lesions were identified as areas lacking normal characteristics (Rabchevsky et al, 2001; Scholpa, 2023).

2.7. Immunofluorescence

Slides containing sections representing the epicenter and 0.5 or 1 mm rostral and caudal were allowed to dry for 1 hour at room temperature. Slides were then rehydrated in 1x phosphate buffered saline for 5 minutes, after which antigen retrieval was conducted using sodium citrate buffer with 0.05% Tween20. Slides were then washed in phosphate buffered solution with 0.05% Tween 20 (PBST) and blocked using goat serum in PBST at room temperature for 1 h. Primary antibodies for Iba1 (1:500) and GFAP (1:500) were applied and the slides placed in a humidified chamber overnight at 4 °C. Slides were then washed using PBST, and secondary antibody was made up in 5% goat serum in PBST and applied for 2 hours in a humidified chamber, protected from light. DAPI (ThermoFisher, D21490) was used as a nuclear counterstain. Fluorescently labeled spinal cord tissue sections were imaged using an EVOS 10x fluorite LWD phase-contract 0.45NA/6.12WD objective on the Invitrogen EVOS-M5000 imaging platform. At least 3 regions of interest were captured per section. Corrected total fluorescence of regions of interest were quantified using ImageJ. Scale bar is 200 μm.

2.8. Locomotor capability

Locomotor capability was assessed using the 10-point (0–9) Basso Mouse Scale (BMS) (Basso et al, 2006) by a blinded observer. Each mouse was observed for approximately 3 minutes, with bladder expression performed prior to assessment. Animals were observed 24 hours after surgery and weekly thereafter until euthanasia. As expected, sham animals maintained a BMS score of 9 throughout the experiments.

2.9. Thermal hyperalgesia

Thermal hyperalgesia was assessed using the Hargreaves apparatus (Ugo Basile) to assess sensitivity to a mild thermal stimulus (Cheah et al, 2017; Scholpa, 2023). Briefly, mice were allowed to acclimate for at least 30 minutes in the apparatus prior to assessment. The provided heat source with infrared intensity of 50 was applied to the middle of the hind paw and the response latency was automatically recorded via the detector. A maximum latency of 20 seconds was used, at which point the heat source automatically shut off to avoid tissue damage. Each hind paw was assessed 5 times with a 5-minute recovery between each test. The highest and lowest value for each paw was discarded and the data for both paws averaged. A baseline thermal responsiveness was obtained one day prior to SCI and mice were assessed weekly beginning 21 DPI until euthanasia. Assessment was done prior to that day’s drug administration.

2.10. Assessment of mechanical allodynia

Mechanical allodynia was assessed using the von Frey apparatus and filaments (Vanderah et al, 2000). Mice were acclimated to the apparatus for 1 hour and to the filaments/action of taking measurements over the course of 4 days prior to baseline measurements. The filaments used were 2.44 (0.04 g), 2.83 (0.07 g), 3.22 (0.16 g), 3.61 (0.4 g), 4.08 (1 g), and 4.56 (4 g). On testing days, manual bladder expression was performed, and mice were placed in the apparatus for at least 1 hour prior to assessment. Starting with the 3.61 filament, each was applied perpendicularly to the plantar surface of the hind paws in the “up-and-down” method, whereby the paw threshold was determined by using either increasing or decreasing tinsel strength filaments based on the response of the animal. Measurements were taken until either maximum or minimum filament strength was used, or 4 measurements after the first positive withdrawal. Data for both paws were averaged. A baseline response was obtained one day before SCI and mice were assessed weekly beginning 21 DPI until euthanasia. Assessment was done prior to that day’s drug administration.

2.11. Statistical analysis

Data are expressed as mean ± SEM Investigators conducting behavioral experiments were blinded to treatment. Tissue isolated from a single animal or a single animal’s behavior represents an n = 1. Sample sizes of experiments were similar to those we published previously (Scholpa et al, 2019a,b, 2021; Simmons et al, 2020, 2021; Scholpa, 2023). Total n for each group can be found in the figure legends. Animals that exhibited self-mutilation were killed and excluded from the study as neither functional capability nor pain-related behavior could be accurately measured. No self-mutilation occurred in the moderate injury groups, and 2 mice were excluded for self-mutilation in the SCI + Vehicle and SCI + Formoterol severely injured groups. Behavior and body weight data were analyzed using two-way ANOVA with repeated measures followed by Dunnett’s post hoc test. Differences between 2 groups were analyzed using two-tailed Student’s t test, while that of 3 groups or more was analyzed using one-way ANOVA followed by Tukey’s post hoc test. In all cases, GraphPad Prism software was used, and a P value of .05 or less was considered indicative of a statistically significant difference between means.

3. Results

3.1. Formoterol treatment alters inflammatory-related gene expression in the injury site 24 hours after severe SCI

Female mice were subjected to an 80 kdyn force-controlled impactor-induced contusion model of SCI followed by treatment with vehicle or 0.3 mg/kg formoterol 8 hours after injury. The injury site was extracted 24 hours post-SCI and flash frozen in liquid nitrogen. TempO-Seq, a high-throughput microarray tool designed by BioSpyder, followed by Gene Ontology Biological Process (GOBP) analysis was used to determine specific gene expression differences across groups. Identification of networks and clusters of altered pathways revealed that formoterol-treated mice exhibited decreases in immune response and inflammatory-related processes, such as somatic immunity, leukocyte tethering/rolling, chemokine production and monocyte chemotaxis (Fig. 1, Supplemental Table 1). The normalized enrichment score (NES) of specific GOBPs in each cluster are identified in Table 2.

Fig. 1.

Effect of formoterol on inflammatory-related GOBPs in the injury site 24 hours after severe SCI. Female mice were subjected to an 80 kdyn force-controlled impactor-induced contusion model of SCI followed by treatment with vehicle or formoterol (0.3 mg/kg) 8 hours after injury. TempO-Seq followed by gene set enrichment analysis was used to identify networks of significantly different GOBPs. The inflammatory-related clusters of somatic immunity (A), leukocyte tethering/rolling (B), chemokine production (C), and monocyte chemotaxis (D) were altered in the injury site of formoterol- vs vehicle-treated mice 24 hours post-SCI. GOBPs represented by blue circles were decreased with formoterol treatment, while red represents an increase. Data are representative of 6 mice per group.

Table 2.

NES of specific inflammatory-related GOBPs in the injury site of formoterol- vs vehicle-treated mice 24 hours and 7 days after severe SCI

Negative numbers represent formoterol-induced decreases; positive numbers represent increases.

	NES	Somatic Immunity
1	−2.174	GOBP_POSITIVE_REGULATION_OF_B_CELL_MEDIATED_IMMUNITY
2	−2.107	GOBP_POSITIVE_REGULATION_OF_HUMORAL_IMMUNE_RESPONSE_MEDIATED_BY_CIRCULATING_IMMUNOGLOBULIN
3	−2.061	GOBP_IMMUNOGLOBULIN_PRODUCTION_INVOLVED_IN_IMMUNOGLOBULIN_MEDIATED_IMMUNE_RESPONSE
4	−2.003	GOBP_SOMATIC_DIVERSIFICATION_OF_IMMUNOGLOBULINS_INVOLVED_IN_IMMUNE_RESPONSE
5	−1.994	GOBP_POSITIVE_REGULATION_OF_NATURAL_KILLER_CELL_MEDIATED_IMMUNITY
6	−1.994	GOBP_POSITIVE_REGULATION_OF_NATURAL_KILLER_CELL_MEDIATED_CYTOTOXICITY
8	−1.75	GOBP_REGULATION_OF_ISOTYPE_SWITCHING
7	−1.738	GOBP_SOMATIC_DIVERSIFICATION_OF_IMMUNE_RECEPTORS_VIA_SOMATIC_MUTATION
9	−1.696	GOBP_POSITIVE_REGULATION_OF_TYPE_2_IMMUNE_RESPONSE
10	−1.647	GOBP_REGULATION_OF_CELL_KILLING
11	−1.601	GOBP_SOMATIC_DIVERSIFICATION_OF_IMMUNOGLOBULINS
12	−1.592	GOBP_POSITIVE_REGULATION_OF_KILLING_OF_CELLS_OF_ANOTHER_ORGANISM
13	−1.588	GOBP_PLASMA_CELL_DIFFERENTIATION
14	−1.569	GOBP_POSITIVE_REGULATION_OF_MYELOID_LEUKOCYTE_MEDIATED_IMMUNITY
15	−1.548	GOBP_POSITIVE_REGULATION_OF_LEUKOCYTE_MEDIATED_IMMUNITY
16	−1.53	GOBP_REGULATION_OF_LEUKOCYTE_MEDIATED_CYTOTOXICITY
17	−1.53	GOBP_REGULATION_OF_LYMPHOCYTE_MEDIATED_IMMUNITY
18	−1.518	GOBP_POSITIVE_REGULATION_OF_NEUTROPHIL_ACTIVATION
19	−1.486	GOBP_POSITIVE_REGULATION_OF_ISOTYPE_SWITCHING_TO_IGA_ISOTYPES
20	−1.481	GOBP_B_CELL_ACTIVATION_INVOLVED_IN_IMMUNE_RESPONSE
21	−1.458	GOBP_POSITIVE_REGULATION_OF_LYMPHOCYTE_MEDIATED_IMMUNITY
22	1.537	GOBP_PEPTIDE_ANTIGEN_ASSEMBLY_WITH_MHC_PROTEIN_COMPLEX
		Leukocyte Tethering/Rolling
23	−2.188	GOBP_REGULATION_OF_LEUKOCYTE_TETHERING_OR_ROLLING
24	−1.934	GOBP_LEUKOCYTE_TETHERING_OR_ROLLING
25	−1.742	GOBP_LEUKOCYTE_ADHESION_TO_VASCULAR_ENDOTHELIAL_CELL
26	−1.61	GOBP_REGULATION_OF_LEUKOCYTE_ADHESION_TO_ARTERIAL_ENDOTHELIAL_CELL
27	−1.61	GOBP_LEUKOCYTE_ADHESION_TO_ARTERIAL_ENDOTHELIAL_CELL
28	−1.594	GOBP_REGULATION_OF_LEUKOCYTE_ADHESION_TO_VASCULAR_ENDOTHELIAL_CELL
29	−1.575	GOBP_POSITIVE_REGULATION_OF_LEUKOCYTE_ADHESION_TO_VASCULAR_ENDOTHELIAL_CELL
30	−1.523	GOBP_POSITIVE_REGULATION_OF_LEUKOCYTE_TETHERING_OR_ROLLING
		Chemokine Production
31	−2.199	GOBP_INTERLEUKIN_23_PRODUCTION
32	−2.156	GOBP_CHEMOKINE_C_X_C_MOTIF_LIGAND_2_PRODUCTION
33	−1.864	GOBP_POSITIVE_REGULATION_OF_CHEMOKINE_C_X_C_MOTIF_LIGAND_2_PRODUCTION
34	−1.757	GOBP_POSITIVE_REGULATION_OF_INTERLEUKIN_23_PRODUCTION
35	−1.604	GOBP_POSITIVE_REGULATION_OF_INTERLEUKIN_6_PRODUCTION
36	−1.537	GOBP_POSITIVE_REGULATION_OF_TUMOR_NECROSIS_FACTOR_SUPERFAMILY_CYTOKINE_PRODUCTION
37	−1.535	GOBP_POSITIVE_REGULATION_OF_MONOCYTE_CHEMOTACTIC_PROTEIN_1_PRODUCTION
38	−1.454	GOBP_POSITIVE_REGULATION_OF_CHEMOKINE_PRODUCTION
		Thymocyte/Monocyte Chemotaxis
39	−2.066	GOBP_REGULATION_OF_MONOCYTE_EXTRAVASATION
40	−1.985	GOBP_THYMOCYTE_MIGRATION
41	−1.643	GOBP_REGULATION_OF_LYMPHOCYTE_CHEMOTAXIS
42	−1.565	GOBP_NATURAL_KILLER_CELL_CHEMOTAXIS
43	−1.534	GOBP_POSITIVE_REGULATION_OF_NEUTROPHIL_MIGRATION
44	−1.493	GOBP_POSITIVE_REGULATION_OF_LEUKOCYTE_MIGRATION
45	−1.48	GOBP_POSITIVE_REGULATION_OF_MONOCYTE_CHEMOTAXIS
46	−1.447	GOBP_POSITIVE_REGULATION_OF_T_CELL_MIGRATION

3.2. Formoterol treatment alters cytokine/chemokine protein expression in the injured cord 24 hours after severe SCI

Cytokine/chemokine expression was assessed in the injury and peri-injury site 24 hours post-SCI using a Milliplex magnetic bead panel. Injured mice treated with a single dose of formoterol displayed altered cytokine/chemokine expression in the injury and peri-injury sites (Figs. 2 and 3). The injury site of formoterol-treated mice depicted decreased expression of interleukin (IL)-2 (Fig. 2A; F(2,15) = 4.571, P = .0301, ANOVA), a cytokine known to promote inflammatory responses after injury (Hoyer et al, 2008), and macrophage inflammatory protein 1 α (MIP-1α, Fig. 2B; F(2,15) = 9.193, P = .0025, ANOVA), a proinflammatory chemokine known to recruit inflammatory cells and contribute to secondary damage after SCI (Bhavsar et al, 2015; Pelisch et al, 2020). While all injured mice regardless of treatment displayed increased interferon gamma-induced protein 10 (IP-10) in the injury site, a proinflammatory chemokine known to be increased following acute SCI (Gotsch et al, 2007; Qiao et al, 2022), formoterol-treated mice exhibited decreased expression compared with vehicle-treated mice (Fig. 2C; F(2,15) = 31.70, P < .0001, ANOVA). Formoterol treatment also resulted in increased G-CSF in the injury site (Fig. 2D F(2,15) = 9.449, P = .0022, ANOVA), an anti-inflammatory molecule that counteracts secondary damage and improves functional outcome in SCI patients (Khorasanizadeh et al, 2017; Aschauer-Wallner et al, 2021; Koda et al, 2021).

Table 3.

Cytokine/chemokine protein expression in the injury (INJ) and peri-injury (PI) site 24 hours after severe SCI

		INJ			PI
		Sham	SCI + Veh	SCI + Form	Sham	SCI + Veh	SCI + Form
Cytokine/Chemokine Expression (pg/mL ± SEM)	G-CSF	42.35 ± 8.19	1077.25 ± 273.60	1540.50 ± 493.77∗	46.15 ± 13.48	736.30 ± 335.20	788.7 ± 460.10
	GM-CSF	6.90 ± 0.57	8.53 ± 1.04	9.39 ± 1.87	10.48 ± 1.09	10.18 ± 0.69	10.12 ± 1.20
	IFNg	20.14 ± 1.45	23.06 ± 3.80	25.14 ± 4.51	23.77 ± 3.18	17.22 ± 3.22	18.15 ± 1.69
	IL-1a	26.17 ± 2.96	30.94 ± 3.39	25.46 ± 1.91	38.83 ± 3.03	32.49 ± 1.46	33.08 ± 2.92
	IL-1B	7.56 ± 0.56	3.50 ± 0.26∗	2.35 ± 0.42∗	4.75 ± 1.08	3.61 ± 0.57	2.68 ± 0.56
	IL-2	45.77 ± 2.62	43.74 ± 7.55	27.90 ± 3.55∗	34.59 ± 4.36	25.97 ± 2.27	24.40 ± 2.84
	IL-5	60.12 ± 12.06	46.38 ± 3.47	56.54 ± 8.46	113.80 ± 11.29	100.66 ± 10.01	80.21 ± 8.66
	IL-6	16.99 ± 2.68	221.18 ± 61.39∗	246.16 ± 37.70∗	14.53 ± 2.43	22.67 ± 2.39	28.26 ± 5.38
	IL-9	2734.83 ± 282.82	ND	668.73 ± 116.77∗	5608.50 ± 390.84	3215.50 ± 313.12∗	4001.58 ± 487.22∗
	IL-10	24.74 ± 1.89	20.80 ± 3.23	19.77 ± 1.83	74.36 ± 13.94	59.59 ± 17.38	29.91 ± 6.49
	IL-12 (p70)	3.82 ± 0.37	4.47 ± 0.48	4.05 ± 0.44	4.30 ± 0.49	3.89 ± 0.39	3.92 ± 0.34
	IL-15	34.28 ± 3.85	35.00 ± 2.79	26.79 ± 3.95	50.00 ± 6.44	42.98 ± 3.82	35.59 ± 4.14
	IL-17	6.90 ± 0.70	7.50 ± 0.72	8.66 ± 0.68	5.30 ± 0.76	6.68 ± 0.46	7.16 ± 0.56
	IP-10	33.21 ± 6.70	261.36 ± 29.80∗	147.82 ± 24.64∗#	36.62 ± 6.06	120.26 ± 9.60∗	124.59 ± 21.36∗
	KC	3.75 ± 0.25	19.61 ± 3.68∗	24.28 ± 2.38∗	6.77 ± 1.01	9.68 ± 3.24	10.61 ± 2.73
	MCP-1	13.58 ± 2.72	53.95 ± 5.63∗	47.47 ± 4.28∗	49.82 ± 5.22	35.64 ± 6.66	35.13 ± 5.61
	MIP-1α	50.42 ± 4.63	37.64 ± 2.77	29.25 ± 2.82∗	48.50 ± 2.29	48.84 ± 3.34	46.30 ± 2.92
	MIP-1β	4.92 ± 1.08	31.15 ± 7.69∗	24.43 ± 5.88	7.75 ± 1.49	76.81 ± 12.49∗	24.60 ± 5.13#
	MIP-2	15.11 ± 2.56	35.33 ± 4.47	45.57 ± 8.14∗	22.06 ± 1.77	24.30 ± 3.46	22.70 ± 2.85
	RANTES	ND	ND	ND	1.56 ± 0.18	2.29 ± 0.31	1.86 ± 0.14
	TNFα	2.48 ± 0.39	1.88 ± 0.13	1.91 ± 0.24	4.12 ± 0.36	3.46 ± 0.18	3.42 ± 0.56

ND, not detected in 2 or more samples. IL-4, IL-7, IL-12 (p40), and IL-13 are not included in the table as they were not detected in at least 2 samples in all groups. Data are expressed as mean ± SEM and represent 6–8 mice per group (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh by one-way ANOVA followed by Tukey’s post hoc test).

Fig. 2.

Effect of formoterol on specific cytokines/chemokines in the injury site 24 hours after severe SCI. Female mice were subjected to an 80 kdyn force-controlled impactor-induced contusion model of SCI followed by treatment with vehicle or formoterol (0.3 mg/kg) 8 hours after injury. The injury site was collected 24 hours post-SCI and evaluated for cytokine/chemokine protein expression via a Milliplex magnetic bead panel. Data are expressed as mean ± SEM and represent 6 mice per group (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh by one-way ANOVA followed by Tukey’s post hoc test).

Fig. 3.

Effect of formoterol on MIP-1β expression in the injury and peri-injury site 24 hours after severe SCI. Female mice were subjected to an 80 kdyn force-controlled impactor-induced contusion model of SCI followed by treatment with vehicle or formoterol (0.3 mg/kg) 8 hours after injury. The injury (A) and peri-injury site (B) was collected 24 hours post-SCI and evaluated for cytokine/chemokine protein expression via a Milliplex magnetic bead panel. Data are expressed as mean ± SEM and represent 6 mice per group (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh by one-way ANOVA followed by Tukey’s post hoc test). Increased MIP-1β, a proinflammatory chemokine that participates in the induction of neuropathic pain following nerve injury (Saika et al, 2012), was observed in both the injury (F(2,15) = 5.869, P = .0131, ANOVA) and peri-injury site (F(2,15) = 21.09, P < .0001, ANOVA) of vehicle-, but not formoterol-treated mice (Fig. 3). The complete list and expression data of all cytokines/chemokines assessed can be found in Table 3.

3.3. Formoterol treatment improves locomotor capability and body composition after severe and moderate SCI

Following severe SCI, vehicle-treated mice do not recover function beyond ankle movement, as indicated by a Basso Mouse Scale (BMS) score of approximately 2 (Vekaria et al, 2020; Scholpa et al, 2021). To examine the effect of formoterol on pain-related behaviors following SCI, however, injured mice must be capable of plantar placement at the time of assessment, which equates to a BMS score of ≥3, meaning a less severe injury is required. Female mice were subjected to either an 80 kdyn severe or 60 kdyn moderate force-controlled impactor-induced contusion model of SCI followed by daily treatment with vehicle or formoterol beginning 8 hours after injury and continuing for 6 weeks. Locomotor capability was assessed via BMS, a 10-point scale ranging from complete paralysis (0) to normal hindlimb function (9) (Basso et al, 2006), beginning 24 hours after injury and continuing weekly. Sham controls maintained normal function for the duration of the study. Comparable to that reported previously (Scholpa et al, 2019a,b, 2021), vehicle-treated 80 kdyn mice reached a maximum BMS score of ∼2, while formoterol-treated 80 kdyn mice reached a BMS score of ∼3.8 (Fig. 4A). Vehicle-treated 60 kdyn mice reached a BMS score of ∼4.3. As observed with severe SCI (F(12,138) = 92.14, P < .0001, ANOVA), formoterol-treated 60 kdyn mice (F(12,216) = 26.54, P < .0001, ANOVA) depicted increased BMS score compared with vehicle-treated mice by 1 week after injury, ultimately reaching a BMS score of ∼5.5 (Fig. 4A).

Fig. 4.

Effect of formoterol on functional recovery and body composition following moderate and severe SCI. Female mice were subjected to moderate (60 kdyn) or severe (80 kdyn) SCI using a force-controlled impactor-induced contusion model followed by daily administration of vehicle or formoterol (0.3 mg/kg, i.p.) beginning 8 hours postinjury and continuing for 6 weeks. Locomotor function was assessed using the Basso Mouse scale beginning 24 hours after injury and weekly thereafter (A, all injured groups are different from sham and from 24 hours at all timepoints, #P < .05 compared with respective SCI + Veh by two-way ANOVA with repeated measures followed by Dunnett’s post hoc test). Body weight was assessed prior to injury, 24 hours post-SCI, and weekly thereafter (B; all injured groups are different from sham at all timepoints, #P < .05 compared with respective SCI + Veh, †P < .05 compared with initial measurement; line indicates both impact severities, bracketed line indicates all injured groups). Data are expressed as mean ± SEM. and represent ≥8 mice per group. Left and right gastrocnemius muscles were extracted and weighed 6 weeks after injury (C). Data are expressed as mean ± SEM and represent 7–8 mice per group. Individual values of BMS score and percent body weight 4 weeks post-SCI can be seen in D and E, respectively (∗P < .05 compared with respective sham, #P < .05 compared with respective SCI + Veh by one-way ANOVA followed by Tukey’s post hoc test).

Mice were weighed prior to surgery (day 0), 24 hours after SCI, and weekly thereafter. While all mice subjected to SCI exhibited approximately 10% weight loss by 24 hours after injury, both 80 kdyn (F(14,126) = 9.785, P < .0001, ANOVA) and 60 kdyn (F(14,259) = 9.031, P < .0001, ANOVA) mice treated with 0.3 mg/kg formoterol displayed increased body weight compared with respective vehicle-treated mice by 2 weeks post-SCI, returning to presurgery weight by 3 weeks (Scholpa et al, 2019a, 2021) (Fig. 4B). Left and right gastrocnemius muscles were isolated and weighed 6 weeks after injury. The combined muscle weight was normalized to obtain gastrocnemius mass as a percent of total body weight. Sham controls exhibited a muscle mass of 0.95%–0.98% body weight, consistent with that reported previously in female mice (Graham et al, 2016; Scholpa et al, 2019a, 2021). Gastrocnemius mass was decreased ∼25% post-SCI in both 80 kdyn (F(2,21) = 17.53, P < .0001, ANOVA) and 60 kdyn (F(2,18) = 18.22, P < .0001, ANOVA) vehicle-treated mice but was not decreased in formoterol-treated mice (Fig. 4C). BMS score and body weight data 4 weeks postinjury can be found in Fig. 4, D–E (D – 80 kdyn: F(2,20) = 693.6, P < .0001, ANOVA; 60 dkyn: F(2,43) = 155.7, P < .0001, ANOVA; E – 80 kdyn: F(2,21) = 18.51, P < .0001, ANOVA; 60 kdyn: F(2,39) = 9.208, P = .0005, ANOVA).

3.4. Formoterol treatment alters cytokine/chemokine protein expression in the injured cord 24 hours after moderate SCI

To determine if the acute effects of formoterol on the inflammatory response following SCI are dependent on the severity of the injury, female mice were subjected to either a severe 80 kdyn or moderate 60 kdyn model of SCI, followed by treatment with vehicle or 0.3 mg/kg formoterol 8 hours after injury. The injury site was extracted 24 hours post-SCI and analyzed for cytokine/chemokine expression via immunoblot analysis. Consistent with what was observed with the magnetic bead panel, formoterol treatment increased G-CSF expression in the injury site 24 hours after severe SCI (Fig. 5A; F(2,14) = 4.533, P = .0303, ANOVA). Interestingly, this effect was not observed with moderate SCI (Fig. 5B; F(2,14) = 1.339, P = .2936, ANOVA). Vehicle-, but not formoterol-treated, mice exhibited increased injury site expression of the proinflammatory chemokines MCP-1 (A: F(2,15) = 26.93, P < .0001, ANOVA; B: F(2,14) = 17.35, P = .0002, ANOVA), which is known to play a role in maintaining inflammation after injury (Becker, 2005; Deshmane et al, 2009), and BCA-1 (A: F(2,15) = 10.48, P = .0014, ANOVA; B: F(2,14) = 11.41, P = .0011, ANOVA), regardless of injury severity. NF-ρB, a protein transcription factor that plays a vital role in the induction of various proinflammatory genes (Liu et al, 2017), was increased in the injury site of vehicle-treated mice only after severe SCI (A: F(2,15) = 11.66, P = .0009, ANOVA), yet was decreased in the injury site of formoterol-treated mice regardless of injury severity (B: F(2,14) = 7.374, P = .0065, ANOVA). These data indicate comparable decreases in proinflammatory mediators with formoterol treatment following moderate or severe SCI. Despite the similarities in data obtained from the cytokine/chemokine kit and immunoblot analysis, it should be noted that these assays were performed using 2 separate cohorts of severely injured SCI mice and rely on inherently different techniques to assess protein expression.

Fig. 5.

Effect of formoterol on cytokine/chemokine expression in the injury site 24 hours following severe or moderate SCI. Female mice were subjected to either a severe (80 kdyn, A) or moderate (60 kdyn, B) force-controlled impactor-induced contusion model of SCI followed by treatment with vehicle or formoterol (0.3 mg/kg) 8 hours after injury. The injury site was collected 24 hours post-SCI and analyzed for protein expression of various cytokines and chemokines via immunoblot analysis. Data are expressed as mean ± SEM and represent 5–6 mice per group (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh by one-way ANOVA followed by Tukey’s post hoc test).

3.5. Formoterol treatment alters astrocyte and microglia activation after severe and moderate SCI

Female mice were subjected to either a severe 80 kdyn or moderate 60 kdyn model of SCI, followed by treatment with vehicle or 0.3 mg/kg formoterol 8 hours after injury. The injury site was extracted 24 hours post-SCI and protein expression of astrocyte and microglia markers was assessed via immunoblot analysis. Following both severe (F(2,15) = 18.68, P < .0001, ANOVA) and moderate SCI (F(2,14) = 7.097, P = .0074, ANOVA), all injured mice regardless of treatment exhibited increased expression of the astrocyte marker GFAP, an intermediate filament protein positively correlated with astrocyte reactivity (Yang and Wang, 2015), and the microglial marker Iba1. Interestingly, formoterol treatment further increased GFAP expression, yet decreased Iba1 expression compared with vehicle treatment following severe (F(2,15) = 20.15, P < .0001, ANOVA), but not moderate SCI (F(2,14) = 18.02, P = .0001, ANOVA). Expression of GLT-1, which is responsible for removing excess glutamate from synapses and maintaining central nervous system (CNS) glutamate homeostasis (Sattler et al, 2013; Peterson and Binder, 2019), was increased in the injury site of vehicle-, but not formoterol-treated, mice after severe SCI (F(2,15) = 12.60, P = .0006, ANOVA). Following moderate SCI, no effect on GLT-1 was observed with vehicle treatment, but expression was decreased with formoterol treatment (F(2,14) = 5.403, P = .0182, ANOVA). TMEM119, which is positively correlated with neuroinflammation (Ruan and Elyaman, 2022), was also increased in the injury site of vehicle-, but not formoterol-treated, mice after severe SCI (Fig. 6 A; F(2,15) = 12.55, P = .0006, ANOVA), yet there was no effect observed following moderate SCI (Fig. 6B; F(2,14) = 2.243, P = .1429, ANOVA).

Fig. 6.

Effect of formoterol on astrocyte and microglia markers in the injury site 24 hours following severe or moderate SCI. Female mice were subjected to either a severe (80 kdyn, A) or moderate (60 kdyn, B) force-controlled impactor-induced contusion model of SCI followed by treatment with vehicle or formoterol (0.3 mg/kg) 8 hours after injury. The injury site was collected 24 hours post-SCI and analyzed for protein expression of astrocyte and microglia markers via immunoblot analysis. Data are expressed as mean ± SEM and represent 5–6 mice per group (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh by one-way ANOVA followed by Tukey’s post hoc test).

Female mice were subjected to either a severe 80 kdyn or moderate 60 kdyn model of SCI, followed by daily treatment with vehicle or 0.3 mg/kg formoterol beginning 8 hours after injury and continuing for 7 d. The epicenter, as well as sections 0.5 and 1 mm caudal and rostral were stained with the astrocyte marker GFAP and microglia marker Iba1 (Fig. 7A and D). While no difference was observed with treatment following severe SCI (Fig. 7B; F(4,70) = 0.09750, P = .9829, ANOVA), there was an increase in GFAP positive staining in the epicenter of formoterol- versus vehicle-treated mice 7 days after moderate SCI (Fig. 7E; F(4,55) = 1.438, P = .2337, ANOVA). No difference in Iba1 was observed with either injury severity (Fig. 7C and F: C: F(4,70) = 0.2863, P = .8859, ANOVA; F: F(4,55) = 0.4753, P = .7536, ANOVA).

Fig. 7.

Effect of formoterol on astrocyte and microglia infiltration 7 days following severe or moderate SCI. Female mice were subjected to either a severe (80 kdyn, A-C) or moderate (60 kdyn, D–F) force-controlled impactor-induced contusion model of SCI followed by daily treatment with vehicle or formoterol (0.3 mg/kg) beginning 8 hours after injury and continuing for 7 days. The spinal cords were extracted and the epicenter, as well as sections 0.5 and 1 mm caudal and rostral to the epicenter were stained with GFAP (green), Iba1 (red), and DAPI nuclear stain (blue). Data are expressed as mean ± SEM and represent 6 mice per group (∗P < .05 compared with SCI + Veh at respective location by Student’s t test). Scale bar is equal to 200 μm.

3.6. Formoterol treatment decreases lesion volume 7 days after severe and moderate SCI

Lesion volume was assessed using Eriochrome cyanine staining for myelin 7 days following severe or moderate SCI (Fig. 8, A and E). Similar to previous reports (Scholpa et al, 2019b, 2021), analysis revealed decreased lesion volume (Fig. 8B; t = 2.532, P = .0298, unpaired t test) and increased percent spared tissue (Fig. 8C; t = 3.282, P = .0083, unpaired t test) in formoterol- compared with vehicle-treated mice following severe SCI. Similar effects were observed following moderate SCI (Fig. 8, F and G: F: t = 2.537, P = .0349, unpaired t test; G: t = 6.203, P = .0003, unpaired t test). Importantly, total volume analyzed was not different between treatment groups for either injury severity (Fig. 8, D and H: D: t = 0.06313, P = .9509, unpaired t test; H: t = 0.7336, P = .4841, unpaired t test).

Fig. 8.

Effect of formoterol on lesion volume 7 days following severe or moderate SCI. Female mice were subjected to either a severe (80 kdyn, A–D) or moderate (60 kdyn, E–H) force-controlled impactor-induced contusion model of SCI followed by daily treatment with vehicle or formoterol (0.3 mg/kg) beginning 8 hours after injury and continuing for 7 days. The spinal cords were extracted, stained with Eriochrome cyanine (A, E) and analyzed for lesion volume (B, F), percent spared tissue (C, G), and total volume (D, H) across 2 mm of cord. Data are expressed as mean ± SEM and represent 5–6 mice per group (∗P < .05 compared with SCI + Veh by Student’s t test).

3.7. formoterol treatment decreases pain-related behaviors after moderate SCI

Response to thermal (Fig. 9A) and mechanical (Fig. 9B) stimuli were assessed prior to injury and weekly beginning 3 weeks after moderate SCI. As expected, no difference in withdrawal response was observed in sham mice at any time point for either assessment. All injured mice depicted a decrease in withdrawal latency of ∼40% following thermal stimulation by 3 weeks after injury, indicative of thermal hyperalgesia. Although they did not recover to pre-SCI response, formoterol-treated mice exhibited significant reversal of SCI-induced thermal hypersensitivity compared with vehicle-treated mice 5 and 6 weeks post-SCI (Fig. 9A; F(8,144) = 6.110, P < .0001, ANOVA). In contrast, while vehicle-treated mice displayed persistent mechanical allodynia, as evidenced by ∼60% decrease in withdrawal time beginning 3 weeks after injury and continuing through the duration of the study, formoterol-treated mice did not exhibit altered withdrawal response to static mechanical stimulation at any timepoint assessed (Fig. 9B; F(8,116) = 13.14, P < .0001, ANOVA). Withdrawal response 6 weeks postinjury can be found in Fig. 9, C–D (C: F(2,35) = 34.09, P < .0001, ANOVA; D: F(2,29) = 114.5, P < .0001, ANOVA). Data are presented as Percent Pre-SCI to account for basal differences between mice and groups, and to allow for analysis of the effect of injury and treatment within each individual animal. Analysis of unnormalized raw thermal hyperalgesia data (Fig. 9E) depicts decreased withdrawal time in both injured groups at all timepoints, with a trending increase in latency in the formoterol- compared with vehicle-treated group in weeks 5 and 6 (F(8,144) = 7.892, P < .0001, ANOVA). Analysis of unnormalized mechanical allodynia data (Fig. 9F) results in similar findings to that of the normalized data, with formoterol-treated mice exhibiting no allodynia at any timepoint (F(8,116) = 10.78, P < .0001, ANOVA).

Fig. 9.

Effect of formoterol on thermal hyperalgesia and mechanical allodynia following moderate SCI. Female mice were subjected to a moderate (60 kdyn) force-controlled impactor-induced contusion model of SCI followed by daily treatment with vehicle or formoterol (0.3 mg/kg) beginning 8 hours after injury and continuing for 6 weeks. Thermal hyperalgesia (A) and mechanical allodynia (B) were assessed in separate cohorts via Hargreaves apparatus and von Frey filaments, respectively, prior to injury and weekly beginning 3 weeks after SCI and data are expressed as Percent Pre-SCI for each animal (∗P < .05 compared with sham, #P < 0.05 compared with SCI + Veh, †P < .05 compared with Pre-SCI by two-way ANOVA with repeated measures followed by Dunnett’s post hoc test). Individual values of thermal hyperalgesia and mechanical allodynia 6 weeks post-SCI can be seen in C and D, respectively (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh one-way ANOVA followed by Tukey’s post hoc test). Raw data for these assessments can be found in E and F. Data are based on the average of both hind paws, are expressed as mean ± SEM., and represent ≥8 mice per group (∗P < .05 compared with sham, #P < .05 compared with SCI + Veh, †P < .05 compared with Pre-SCI by two-way ANOVA with repeated measures followed by Dunnett’s post hoc test; line indicates all timepoints, bracketed line indicates all injured groups at all timepoints).

4. Discussion

We have previously shown that treating daily with formoterol, an FDA-approved long-lasting ADRB2 agonist, induces MB and improves recovery post-SCI in mice (Scholpa et al, 2019a,b, 2021; Scholpa, 2023). Despite these promising data, the full extent and mechanisms of formoterol-induced recovery have yet to be determined. Studies have shown that formoterol treatment attenuates neuroinflammation in a mouse model of Alzheimer’s disease (Abdel Rasheed et al, 2018). Additionally, ADRB2 agonism by tricyclic antidepressants leads to anti-inflammatory activity and reduction of neuropathic pain following treatment with chemotherapeutics (Yalcin et al, 2009, 2010). Neuropathic pain is caused by the interruption of ascending and descending nerve fibers and can manifest post-SCI as below-injury, at-injury, or above-injury pain, all of which are largely resistant to pharmacological intervention (Masri and Keller, 2012). Therefore, we hypothesized that formoterol treatment alters the inflammatory response after SCI, contributing to enhanced recovery and reduction in neuropathic pain.

Neuroinflammation is a well known consequence of, and contributor to, injury progression after CNS injury. After SCI, the release of cytokines and chemokines becomes a driving factor in the destructive inflammatory response and glial scar formation brought on by astrogliosis (Kwiecien et al, 2020). We observed downregulation of genes associated with various inflammatory processes with formoterol treatment, including those involved in B-cell mediated immunity, which suggests that formoterol may dampen the B-cell response after SCI, which could protect surrounding tissue from poly-reactivity to antigens that can become autoaggressive, leading to an increased inflammatory response (Wekerle, 2006; Jain and Yong, 2022). Additionally, we observed downregulation of genes associated with natural killer cell–mediated immunity, which could lead to decreased secretion of proinflammatory cytokines, further indicating an anti-inflammatory effect of formoterol (Paul and Lal, 2017). These findings were corroborated by decreased expression of proinflammatory proteins in the injury site of formoterol-treated mice, including IL-2, which has been implicated in the proliferation of natural killer cells and B-cells after SCI (Mukhamedshina et al, 2017), MIP-1α, a chemotactic chemokine involved in the proinflammatory cascade via regulation of activated macrophages and astrocytes after SCI (Pelisch et al, 2020), and IP-10, which is secreted by various cell types, including activated microglia, and has a variety of proinflammatory functions including the secretion of additional chemokines, resulting in a sustained proinflammatory cascade (Qiao et al, 2022). Furthermore, MIP-1β, which promotes breakdown of the blood-CNS barrier after injury (Yang et al, 2019; Estevao et al, 2021) was decreased in formoterol-treated mice. Importantly, upregulation of both MIP-1α and MIP-1β have been associated with pain induction after CNS injury (Kiguchi et al, 2010; Saika et al, 2012). Taken together, these data indicate that formoterol decreases proinflammatory processes after injury, suggesting a neuroprotective effect.

Formoterol treatment also upregulated G-CSF, an anti-inflammatory cytokine that activates anti-inflammatory neutrophils and reduces the release of proinflammatory mediators (Vafaei Mastanabad et al, 2023). Importantly, G-CSF suppresses apoptosis, limits expression of proinflammatory cytokines, increases hindlimb recovery, and plays a neuroprotective and neuroregenerative role after CNS injury, including SCI (Kadota et al, 2012; Aschauer-Wallner et al, 2021). We also observed formoterol-induced upregulation of genes involved in antigen assembly with the major histocompatibility complex. Major histocompatibility complex 1 is hypothesized to be involved in the maintenance of inhibitory synapses following injury, leading to a shift in cellular efforts toward regeneration versus firing action potentials (Oliveira et al, 2004; Thams et al, 2008). Shifting efforts toward regeneration of injured neurons would be invaluable to recovery after SCI and could aid in the reduction of neuropathic pain after injury (Wei et al, 2022). Increased regeneration of CNS tissue coupled with a more anti-inflammatory environment could be contributing to formoterol-induced recovery.

Astrocytes and microglia play key roles in the inflammatory response, having both negative and positive consequences after injury. While integral to neuroinflammation and glial scar formation, previous studies have shown that ablation of reactive astrocytes or depletion of microglia after SCI can have detrimental effects on recovery (Faulkner et al, 2004; Gu et al, 2019; Zhou et al, 2023). Formoterol treatment appeared to have more of an effect on astrocyte and microglia protein markers 24 hours following severe, compared with moderate SCI, increasing GFAP, while decreasing Iba1 and TMEM119. While the reason for this remains to be discerned, it is important to note that these markers are hypothesized to be general astrocyte and microglial markers and do not differentiate between proinflammatory and anti-inflammatory cells (Liddelow and Barres, 2017; Waller et al, 2019). In contrast, formoterol treatment decreased GLT-1 regardless of injury severity. While this could be indicative of a more excitotoxic environment (Pajarillo et al, 2019), this effect could also be in response to the decrease in microglial activation observed with formoterol, as activated microglia are known to increase astrocyte GLT-1 expression (Tilleux et al, 2009). Interestingly, immunofluorescence revealed no formoterol-induced effect on GFAP following severe injury or Iba1 with either injury model by 7 days postinjury, suggesting that the effect of formoterol on glial cell activation and neuroinflammation may occur exclusively in the acute phase following SCI. While increased GFAP positive staining was observed with formoterol 7 days after moderate SCI, this could be attributed to the increase in spared tissue observed with treatment as opposed to a direct effect on astrocyte activation/infiltration. Future studies will examine these effects.

Proinflammatory chemokines and cytokines are known to directly activate nociceptors and induce neuropathic pain (Matsuda et al, 2019). Recent research has also shown that mitochondrial dysfunction induced by chemotherapeutic drugs is correlated with increased incidence of neuropathic pain (Doyle and Salvemini, 2021). Treatment with ADRB2 agonists, including formoterol, are reported to decrease paclitaxel-induced neuropathic pain (Chen et al, 2021), as well as allodynia and hypersensitivity following nerve injury, presumably via modulation of proinflammatory signaling in microglia (Damo et al, 2023). Although we observed formoterol-induced reduction in both thermal hyperalgesia and mechanical allodynia, the effect was more pronounced with mechanical allodynia, with formoterol-treated mice responding similarly to uninjured mice. While both assessments are measures of neuropathic pain, the pathways differ. Thermal hyperalgesia is mediated by heat-nociceptive C-fibers and intended to be protective (Shir and Seltzer, 1990; Wu and Ringkamp, 2020), while mechanical allodynia has been associated with A-beta, A-delta, and mechanosensitive C-fibers (Basbaum et al, 2009; Xu et al, 2015). The mechanism behind this disparity in nociceptive pain-type amelioration with formoterol treatment will be further explored in future work. Nonetheless, the formoterol-induced decrease in both thermal hyperalgesia and mechanical allodynia after SCI support existing work implicating ADRB2 agonism in the reduction of neuropathic pain.

In general, female rodents are preferred for the study of SCI due to the relative ease of manual bladder expression after injury decreasing the likelihood of developing urinary tract infections (Onifer et al, 2007). However, nearly 80% of SCI cases occur in males (National Spinal Cord Injury Statistical Center, 2022). We have previously shown that daily 0.3 mg/kg formoterol treatment beginning 8 hours after SCI induces MB and improves functional recovery, lesion volume and body composition in male mice similar to that in females (Scholpa et al, 2021). While we would anticipate comparable formoterol-induced effects in male mice as shown here in female mice, we acknowledge that sex can contribute to both altered pharmacological responses (Anderson, 2008) and pain response, with sex hormones having been linked to chronic pain (Lenert et al, 2021; Osborne and Davis, 2022). Interestingly, human and animal studies have also identified sex-specific distinctions in pain mechanisms, including genetic and immune system differences, that contribute to altered pain sensitivity (Mogil, 2020). Therefore, the effect of formoterol on pain and chemokines/cytokines in males will be explored in future studies.

It is important to note that the data presented here is focused on stimuli-evoked pain-related behaviors, while clinically, neuropathic pain post-SCI typically presents as spontaneous (Shiao and Lee-Kubli, 2018). As such, future work will also evaluate spontaneous pain-related behaviors with formoterol treatment via flinching and/or guarding assays (Thompson et al, 2020). Similarly, the effect of formoterol on nonreflexive pain, such as weight bearing and voluntary movement (Cho et al, 2013; Griffioen et al, 2015; Quadros et al, 2015; Hasriadi et al, 2021; Álvarez-Pérez et al, 2022), has not yet been assessed. Unfortunately, many techniques for assessing nonreflexive pain are dependent on locomotor function, which could confound data interpretation with this model.

We have previously shown that formoterol treatment improves multiple downstream aspects of SCI, including enhanced mitochondrial function and body composition, as well as decreased spread of injury (Scholpa et al, 2019a,b). As such, it is plausible that multiple, if not all, of these effects could be contributing to the decreased pain-related behaviors observed. In addition to MB induction, formoterol ameliorates neuropathic pain-related behaviors mediated by peripheral delta opioid receptors and microglia-dependent ADBR2 expression (Ceredig et al, 2019; Kremer et al, 2020; Chen et al, 2021; Damo et al, 2023). Therefore, these pathways could also be affecting the results seen here. Similarly, given the rising evidence of the role of dorsal root ganglia stimulation for the alleviation of chronic pain (Berger et al, 2021), the effect of formoterol on these structures would be worth investigation. The reflexive pain behaviors assessed are mediated by lower lumbar spinal cord processes, while the analyses performed herein were exclusive to the injury and peri-injury regions (Ma and Eisenach, 2003). Therefore, future work aimed at uncovering the mechanisms by which formoterol treatment improves not only functional recovery, but also pain-related behaviors after SCI, will include investigation of the lumbar spinal cord. Taken together, the data presented here suggest that treatment with the FDA-approved ADRB2 agonist formoterol mitigates neuroinflammation and neuropathic pain post-SCI, further reinforcing its potential as a powerful alternative to current treatment strategies.

Conflict of interest

The authors declare no conflicts of interest.