Use of live animals was conducted in accordance with PHS Policy on Humane Care and Use of Laboratory Animals and the Guide and was approved by the Institutional Animal Care and Use Committee at James J. Peters Veterans Affairs Medical Center (JJP VAMC) IACUC #CAR-20-11. Male and female C57Bl6 mice were purchased from Charles River and housed with controlled photoperiod (12/12 h light/dark cycle) and temperature (23–25°C) with ad libitum access to water and pelleted chow. Dark cycle was set from 6:00 AM to 6:00 PM. Animals were single housed one-week prior surgery. Experiments using cultured astrocytes were performed following protocols approved by Ethics Committee of the Universidad de Valparaíso, Chile; ID: #CICUAL F-03.

A general summary of animal groups and sample sizes are provided in Table 1. At 4 months of age, male and female animals of similar weight (Supplementary Figures 1A, B) were randomly assigned to SCI or a sham-SCI. Each group was then randomly assigned to boldine or vehicle-treated groups (Table 1). The experimental design and timeline for the study procedures is shown in Figure 1.

A motor-incomplete contusion SCI was performed using an Infinite Horizons (IH) impactor (Precision Systems and Instrumentation) as previously reported (Toro et al., 2021). Briefly, after induction of anesthesia by 3% isoflurane inhalation, animals were placed on heating pads warmed with recirculating water to 37°C. Hair was clipped over the cervical and thoracic spine areas. All animals underwent a laminectomy to expose the dura at level of thoracic vertebrae 9 (T9). The size of laminectomy was approximately 2 mm in diameter to allow the probe to impact the dura without touching any bone or other surrounding tissues. Randomized animals selected for the sham-SCI group had the vertebral muscle around the laminectomy site sutured to stabilize the spinal column, the incision site closed with 7 mm wound clips, and were returned to a clean cage on top of a warming pad. For animals selected for the SCI groups, they were placed on the clamping platform of the IH impactor under 3% isoflurane where their vertebral column was stabilized using the attached forceps to the IH clamping platform and received a 65 kdyne contusion SCI (Scheff et al., 2003). We chose this impact force as it results in a moderate-severe injury that allows for a partial recovery function over the timeframe of the study. Animals reach a maximal recovery and then plateau at around 28 days post-injury. After the lesion is generated, the left-right symmetry was confirmed by detection of equal bruising on both sides of the dorsal median sulcus. Actual impact force and cord displacement values were recorded for each animal and group (Supplementary Figure 2). Muscle was closed using resorbable sutures and skin closed with 7-mm wound clips. Post-operative care took place in cages with Alpha-Dri bedding (Newco Distributors, Inc., Hayward, CA, USA) over heating pads warmed with recirculating water for the first 72 h. All animals received pre-warmed lactated Ringers solution (LRS), carprofen and Baytril every 24 h for 5 d. Wound clips were removed at 10 days post-injury (dpi). Urine was expressed manually twice a day by gentle pressure and massage of the bladder until spontaneous voiding. Animals were fully checked for signs of stress, urine scalds, and autophagia at least once a day for the length of the study.

Daily boldine (Millipore Sigma, Jaffrey, NH, USA) administration started at 3 dpi. Boldine was added to cages at 2:00 PM. It was prepared by dissolving boldine in a mix of DMSO and peanut oil (Sigma). This mixture was then added to peanut butter (PB) so that 1.0 g of total bolus could be given once per day at a dose of 50 mg/kg. The final concentration of DMSO was less than 2.5%. Single-housed animals were familiarized with 1.0 gram of peanut butter for a week prior to surgery. All animals consumed 100% of their daily PB mix within 1 h and continued to do so throughout the remainder of the study. Vehicle-treated SCI and laminectomy-only animals (shams) received daily equal amount of the PB mix without boldine.

Locomotor recovery after SCI was tested as previously reported (Toro et al., 2021) using the Basso mouse scale (BMS) open-field test (Basso et al., 2006), and the horizontal ladder rung walk test (LRWT) (Cummings et al., 2007), at specified timepoints (Figure 1). BMS is a well-established method for evaluating the severity of impairments in locomotor function after SCI by scoring locomotor milestones using a 9-point scale, with 0 being completely hindlimb paralysis and 9 being a normal healthy gait. The LRWT is used to evaluate fine motor skills and coordinated stepping as mice attempt to cross a commercially available horizontal ladder and were evaluated as they attempted to cross (Cummings et al., 2007). For BMS, two blinded investigators independently scored the animals and the results were averaged for a final score. For LRWT, animals were recorded with a video camera located under the ladder that was moved manually to keep the animal in-frame and later reviewed by two blinded observers who recorded the number of correct steps and errors. Animals were familiarized with equipment used for each of these behavioral tests one-week prior to surgery. We performed all behavior tests during the animals’ dark phase of the light:dark cycle, which was set from 6:00 AM to 6:00 PM for this study, as animals are normally more active in the dark. BMS and LRWT were performed between 9:30 AM and 11:30 AM.

At days 14 and 28 post SCI, 5 animals per group (Table 1) were randomly selected to undergo perfusion-fixation under a deep anesthesia following an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (30 mg/kg) as previously reported (Toro et al., 2021). Briefly, mice were euthanized by transcardial perfusion with sterile saline followed by injection if ice-cold 4% paraformaldehyde (PFA). After perfusion, spinal cords were removed and further post-fixed in 4% PFA for 72 h, transferred to a solution of 30% sucrose and stored at 4°C until cryostat sectioning. In addition, fresh spinal cord tissues from remaining animals of every group were collected after inducing deep anesthesia by inhalation of 3% isoflurane followed by decapitation. Spinal cord segments (∼2 mm) containing half the lesion epicenter and the region immediately rostral or caudal, were collected and snap frozen in liquid nitrogen and then stored at −80°C for biochemical analysis; tissues from comparable areas of spinal cords of sham-controls were also collected.

Transverse 10-micron sections of perfusion-fixed spinal cords were obtained with a cryostat (Leica, West Hollywood, CA, USA) and used for immunofluorescence staining as previously reported (Toro et al., 2021). Myelin was stained using FluoroMyelin according to the manufacturer’s protocol (FluoroMyelin green, ThermoFisher, Carlsbad, CA, USA). Spinal cords sections from SCI animals from different groups were transversally cut rostral and caudal from the injury site every 100 microns for a total of 2 mm length centered on the injury epicenter. Sections were imaged with a confocal microscope (Carl Zeiss, Jena, Germany) as described below. To visualize differences in spared myelin an image analysis was performed using the ImageJ software as described elsewhere (Streijger et al., 2016). Briefly, thresholds for signals in the images were set at background gray values. The region of interest was determined and mean intensity and total number of pixels above threshold were measured. Spared tissue was determining by the percentage of white matter calculated by the total area of the spinal cord for each section using ImageJ.

Sections of perfusion-fixed spinal cord were used for immunofluorescence staining as previously reported (Toro et al., 2021). Antibodies were used to detect the following proteins: GAP-43 (Abcam #ab16053) to label regenerating axons and identify sprouting; GFAP (Abcam #ab7260) to detect reactive astrocytes; and Iba1(Abcam #ab178846) for labeling macrophages and activated microglia. Secondary antibodies, Alexa Fluor 488-conjugated goat anti-mouse IgG (Abcam # ab150113) and Alexa Fluor 647-conjugated goat anti-rabbit IgG (Abcam # ab150079). Details in Table 2.

Fixed 10-micron spinal cord sections were mounted on Superfrost Plus slides (Thermo Fisher Scientific, Pleasanton, CA, USA). Customized probes for GFAP, Cx43 and S100a8 were designed and provided by Advance Cell Diagnostics (Hayward, CA, USA) for detection of mRNA. In situ hybridizations were performed according to the RNAscope Multiplex Fluorescent Reagent Kit v2 Assay protocol provided by the manufacturer and as described previously (Wang et al., 2012).

5 × 5 tiled images were obtained from stained sections using a 20× objective and a Zeiss 700 confocal microscope (Carl Zeiss, Jena, Germany). Blinded quantification was performed using ImageJ software (version 2.1.0/1.53c, National Institute of Health, USA) and integrated density of pixels was measured for each section and a mean value was calculated as previously described (Oliveira et al., 2004; Toro et al., 2021). Maximum background threshold was determined for each image and set for intensity quantification. Data are represented using the mean ± SEM.

Total RNA was extracted from spinal cord segments that extended either ∼2 mm rostral or ∼2 mm caudal to lesion epicenter using TRIzol reagent (ThermoFisher, Carlsbad, CA, USA) following the manufacturer’s instructions and methods previously described (Toro et al., 2018, 2021). Total RNA concentrations were determined by absorbance at 260 nm using a Nanodrop spectrophotometer (Thermo Scientific, Carlsbad, CA, USA). RNA was reverse-transcribed into cDNA using Omniscript reverse transcriptase (Qiagen, Germantown, MD, USA). PowerUp SYBR Green Master Mix (ThermoFisher, Carlsbad, CA, USA) was used to measure mRNAs of interest by qPCR. Primers were designed with help of the Primer Blast program from NCBI (Table 3). Formation of single SYBR Green-labeled PCR amplicons were verified by running melting curve analysis. Threshold cycles (CTs) for each PCR reaction were identified by using the QuantStudio 12K Flex software. To construct standard curves, serial dilutions were used from 1/2 to 1/512 of a pool of cDNAs generated by mixing equal amounts of cDNA from each sample. The CTs from each sample were referred to the relative standard curve to estimate the mRNA content per sample; the values obtained were then normalized for variations using peptidylprolyl isomerase A (Ppia) mRNA as the normalizing unit.

We used total RNA extracted from each of two spinal cord segments that were collected from male mice at 14- and 28 days post SCI (N = 3 per timepoint), One of extended 2 mm rostral from the injury epicenter and the second 2 mm caudal from the injury epicenter. RNA integrity was checked using the RNA 6000 Nano assay (Agilent, Santa Clara, CA, USA). The sequencing library was prepared with a standard TruSeq RNA Sample Prep Kit v2 protocol (Illumina, San Diego, CA, USA), as described previously (Mariottini et al., 2019; Toro et al., 2021). RNA libraries were sequenced on the Illumina HiSeq 2000 System with 100 nucleotide pair end reads, according to the standard manufacturer’s protocol (Illumina, San Diego, CA, USA). For RNAseq data analysis, Star 2.5.4b and bowtie 2 2.1.0, samtools 0.1.7, Rsubread 2.10.5 and DESeq2 1.36.0 were used for read alignment to the mouse reference genome GRCm38.p6 using the ensemble gene annotation. At least 25 million reads were sequenced for each biological replicate (Supplementary Figure 3A). The percentage of reads that were successfully aligned to the mouse reference genome was between 84 and 89% (Supplementary Figure 3B). Differentially expressed genes (DEGs) were identified based on a maximum adjusted p-value of 10% (Supplementary Table 1). Up- and down-regulated gene sets were further interrogated using enrichR (Chen et al., 2013), with pathway enrichment analysis using Fisher’s exact test and the Gene Ontology Biological Processes 2018 library (Ashburner et al., 2000; Gene Ontology, 2021; Supplementary Table 2). Predicted pathways were ranked by significance. Significance p-values were transformed into -log₁₀(p-values) and visualized as bar diagrams.

Primary cultures of spinal cord astrocytes of 3 days old C57Bl6 mice were prepared using previously described methods (Kerstetter and Miller, 2012). Briefly, dissociated cells were resuspended in 10 ml DMEM with 10% FBS then transferred to a poly-lysine coated flask, place in an incubator at 37°C and 5% CO₂, and allowed to adhere for 24 h. Flasks were then shaken at 200 rpm at 37°C overnight. The following day, non-adherent oligodendrocytes, neurons, and microglial were removed with aspiration and adherent cells were gently washed with DMEM containing 10% FBS, 10 mL of fresh DMEM with 10% FBS was added. Cells were allowed to recover and proliferate for 10 d before final seeding HeLa cells transfected with mouse Cx30 or Cx26 were used (kind donation from Christian Giaume, College de France, Paris, France and Klaus Willecke, Life and Medical Sciences Institute, Molecular Genetics, University of Bonn, Bonn, Germany, respectively). The activity of Cx HCs was evaluated using the dye uptake method described previously (Cea et al., 2013; Cisterna et al., 2020). In brief, cells were plated onto glass coverslips and bathed with Locke’s saline solution (all concentrations in mM: 154 NaCl, 5.4 KCl, 2.3 CaCl₂, 1.5 MgCl₂, 5 HEPES, 5 glucose, and pH 7.4) containing 5 μM DAPI, a molecule that crosses the plasma membrane through large-pore channels, including Cx HCs (Theriault et al., 1997). Since DAPI fluoresces upon its intercalation between DNA nucleotides, time-lapse recordings of fluorescent images were measured at regions of interest every 30 s for 13 min using a Nikon Eclipse Ti inverted microscope (Tokyo, Japan) and NIS-Elements software. The basal fluorescence signal was recorded in cells only in Locke’s saline solution that contained divalent cations. Then, cells were exposed to divalent cation-free solution (DCFS; Krebs buffer without CaCl₂ and MgCl₂), followed by 50 μM boldine and finally 200 μM La³⁺. Time-lapse fluorescence snapshot images were taken every 15 s. DAPI fluorescence was recorded in regions of interest using a Nikon Eclipse Ti microscope (Japan). Mouse P2X₇R cDNA was cloned into pIRES-EGFP as previously described (Koshimizu et al., 1999) (kindly donated by Dr. Claudio Acuña, Instituto de Química y Biología, University of Santiago, Chile) and transiently transfected into HeLa cells to evaluate the activity of P2X₇Rs. Intracellular calcium ion was detected using Fura-2AM, a ratiometric dye. Cells were incubated in Krebs-Ringer solution (all concentrations in mM: 145 NaCl, 5 KCl, 3 CaCl₂, 1 MgCl₂, 5.6 glucose, 10 HEPES-Na, pH7.4) containing FURA2-AM dye (2 μM) for 45 min at room temperature. The calcium signal was then measured using a Nikon Eclipse Ti microscope equipped with epifluorescence illumination, and images were obtained by using a Clara camera (Andor, Abingdon, UK) at 2 wavelengths of (λ) 340 nm and 380 nm, followed by calculating the ratio of fluorescence emission intensity after stimulation with each one of these two wavelengths. The activity of P2X₇Rs was induced with 100 μM benzoyl ATP followed by the addition of 50 μM boldine. All measurements were performed in ∼20 cells per experiment in a total of at least four independent experiments.

Statistical evaluations were performed with one-way or two-way mixed model analysis of variance (ANOVA), as indicated in the section “Results” and figure legends. Post-hoc comparisons were done using Tukey’s multiple comparison test. P-values of less than 0.05 were considered significant. Statistical calculations were performed using Prism 9 software (Graphpad, San Diego, CA, USA). Data are expressed as mean values are expressed as mean ± SEM.

The effect of boldine on functional recovery after a thoracic motor-incomplete contusion SCI was evaluated by the open-field BMS test (Basso et al., 2006), and the horizontal LRWT (Cummings et al., 2007; Figure 2). Boldine or vehicle administration began at 3 dpi. All animals had maximum BMS scores before surgery (p > 0.999; Figures 2A, B). As expected, shams treated with either vehicle or boldine presented maximum BMS scores at all timepoints (Figures 2A, B). SCI animals treated with boldine or vehicle presented similar BMS scores at day 3 (0.62 vs. 0.61; p > 0.05 for males, and 0.14 vs. 0.25; p > 0.05 for females). However, scores for boldine-treated animals were significantly higher at 7 dpi in males when compared to SCI-vehicle mice by more than one-point on the BMS (3.41 vs. 2.29; p < 0.01). Boldine resulted in greater BMS mean scores in female mice compared to vehicle-treated animals at 7 dpi, though this did not reach our statistical threshold (2.25 vs. 1.53; p > 0.05). The differences between boldine- and vehicle-treated animals became greatest for both males and females at 14 dpi (6.56 vs. 4.61; p < 0.0001 for males, and 4.84 vs. 3.53; p < 0.01 for females) and stayed consistently elevated at 21 dpi (7.44 vs. 5.57 for males; p < 0.0001, and 5.89 vs. 4.88; p < 0.05 for females) and 28 dpi (7.94 vs. 6.11; p < 0.0001 for males, and 6.46 vs. 5.00: p < 0.01 for females) (Figures 2A, B). Omnibus effect for BMS was significant between treatment groups [F(2, 27) = 891.1 for males and F(2, 21) = 233.1 for females]. For LWRT, all sham animals, vehicle or boldine-treated, made no step placement errors (p > 0.999; Figures 2C, D). The SCI-vehicle group demonstrated a numerically higher percentage of foot placement errors when compared to the SCI-boldine group at 10 dpi (46.4 vs. 23.5% for males; p > 0.05, and 48.9 vs. 34.8%; p > 0.05 for females). These differences met statistical thresholds for significance for both sexes at 14 dpi (37.4 vs. 14.7%; p < 0.05 for males, and 38.0 vs. 21.6%; p < 0.05 for females), 21 dpi (30.2 vs. 12.0%; p < 0.01 for males, and 30.1 vs. 16.0%; p < 0.01 for females) and 28 dpi (23.3 vs. 8.9%; p < 0.001 for males, and 22.0 vs. 11.2%; p < 0.01 for females) (Figures 2C, D). Omnibus effect for LWRT was significant between treatment SCI-groups for both males [F(2, 27) = 136.3] and females [F(2, 21) = 70.87]. In addition, no significant sex differences were observed in vehicle-treated SCI animals by BMS [sex effect, p > 0.05, F(1, 16) = 4.277] or LRWT [sex effect, p > 0.05, F(1, 16) = 4.277]. However, we observed significant sex-differences between boldine-treated SCI animals by BMS [p < 0.001, F(1, 16) = 18, 26] and LRWT [p < 0.001, F(1, 16) = 9.090]. Altogether, our behavioral data suggests that boldine administration starting 3 d after a 65 kdyne contusion SCI induces significant improvement of gross and fine locomotor function in both male and female mice.

To determine whether the overall size of the lesion was affected by boldine administration at 28 days post SCI, we performed fluorescent myelin detection (FluoroMyelin; Thermo Scientific, Carlsbad, CA, USA) and confocal imaging of spinal cord transverse sections from vehicle or boldine-treated SCI males every 100 μm including the injury epicenter, and locations rostral and caudal to it (Figures 3A, B). Spared white matter was defined as the area stained with FluoroMyelin and the lesion size was measured as the disrupted area in each section. Area of spared white matter was calculated for each section after which an area under the curve was calculated and showed a significant increase in samples from boldine-treated SCI as compared to vehicle-treated SCI animals [F(2.477, 7.430) = 93.85; p < 0.05; Figure 3C]. Conversely, lesion volume was significantly reduced by boldine as compared to vehicle-treated SCI animals [F(2.594, 7.782) = 202.2; p < 0.05; Figure 3D]. Of note, samples collected at 14 dpi also showed an increase in white spared matter from SCI animals treated with boldine (Supplementary Figure 4). These data suggest that boldine reduces the lesion size and increase the abundance of white matter spared after SCI.

We further compared the injured spinal cord of vehicle or boldine treated animals at 14 dpi by immunofluorescence (IF) examination. This timepoint was chosen because it reflects when locomotor recovery showed maximal differences between vehicle and boldine-treated groups. To assess the effect of boldine on gliosis after SCI, we probed sections taken just rostral to the lesion site for glial fibrillary acid protein (GFAP), a marker of reactive astrocytes (Brenner, 2014), and the ionized calcium binding adaptor molecule 1 (Iba1), a marker associated with macrophage and microglia activation (Peng et al., 2005; Miller et al., 2012). Fluorescent signals of both GFAP and Iba1 were significantly reduced by boldine (Figures 4A, B; p < 0.01 for GFAP and p < 0.05 for Iba1). Moreover, to investigate changes in expression of proteins related to neuronal plasticity and axon growth, we performed immunostaining of growth-associated protein 43 (GAP43) (Koshi et al., 2010). We showed that GAP43 fluorescence was significantly higher in sections rostral to the injury site of boldine-treated mice compared to vehicle-treated animals (Figure 4C; p < 0.001). These data suggest that boldine reduces glial reactivity and stimulates plasticity of spared fibers at 14 dpi after incomplete SCI. Of note, we did not detect significant effects in samples obtained at 28 dpi (data not shown).

It is known that boldine blocks Cx43 and Panx1 HCs in the brain (Yi et al., 2017). To demonstrate that boldine blocks HCs present on the surface of spinal cord astrocytes, primary cultures of spinal cord astrocytes were maintained in divalent-cation-free culture media in the presence of DAPI, which is taken up by live cells through HCs (Yi et al., 2017), with or without 50 μM of boldine. Boldine significantly reduced uptake of DAPI during time-lapse imaging (Supplementary Figure 5; p < 0.001), indicating that it blocked open HCs in these cells.

Multiple Cx other than Cx43 are expressed in the central nervous system, including Cx26 and Cx30 (Rash et al., 2001; Nagy et al., 2003). Therefore, we tested if boldine blocks dye uptake in open Cx26 and Cx30 HC using transfected HeLa cells. Dye uptake studies demonstrated that boldine slowed the rate of dye uptake through both Cx26 and Cx30 HCs (Supplementary Figure 6; p < 0.05). Addition of lanthanum ions, a well-stablished Cx HC blocker (Anselmi et al., 2008), did not further reduce the rate of dye uptake, confirming that boldine acted by blocking Cx26 and Cx30 HC. Because the effects of boldine on P2X₇R uptake has not been tested, additional cytoplasmic calcium signal studies were performed. The data showed that 50 μM boldine also blocks P2X₇R (Supplementary Figure 7).

An unbiased understanding of the cellular and molecular mechanisms responsible for the differences observed in locomotor recovery between SCI animals treated with vehicle versus boldine was explored through bulk RNA-seq followed by a bioinformatic analysis. Total RNA was extracted from spinal cord segments rostral and caudal from the lesion epicenter (∼2 mm each) at 14 dpi, time when functional recovery differences were maximal in between treatments, and at 28 dpi, when functional recovery had reached a plateau in maximal locomotor function. Differentially expressed genes (DEG) were identified and the magnitude of change and number of DEGs that differed between boldine and vehicle treated SCI animals was reported (Figure 5 and Supplementary Figure 8). The data showed unique differences within the spinal cord segment at 14 dpi: while we found only 6 and 18 up- and downregulated genes in the spinal cord segment rostral from the lesion in boldine-treated animals (Figure 5A), we identified 426 and 913 up- and downregulated genes (Figure 5B) in the segment caudal from the lesion in boldine treated SCI mice, respectively.

To understand more in depth the biology associated with changes in numbers of DEGs, we identified biological processes represented by up- and downregulated genes rostral and caudal from the lesion site (Figures 5C–F). While the top 15 predicted pathways rostral from the lesion did not contain any pathways related to recovery of synaptic function or neuronal repair for any list of DEGs, the top upregulated pathways caudal from the lesion for boldine treated animals at 14 dpi focused on functions related to neuronal development, synaptic transmission and axonogenesis (Figure 5D). In contrast, the pathways that were downregulated caudal from the lesion included functions such as catabolic and biosynthetic processes (Figure 5F).

To further investigate the effect of boldine in changes of mRNA levels after SCI, differences in targeted gene expression between vehicle and boldine treated animals were evaluated using spinal cord segments rostral and caudal from the lesion site at 14 dpi. RT-qPCR experimentation revealed significantly higher levels of pro-inflammatory molecules Ccl2 (Figure 6A; p < 0.0001), IL-6 (Figure 6B; p < 0.001) and S100a (Figure 6C; p < 0.001) in samples from SCI animals treated with vehicle as compared to shams. Interestingly, these changes were not detected in boldine-treated SCI samples, and were, in fact, significant between vehicle and boldine-treated SCI animals (Figures 6A–C; p < 0.05). Moreover, similar trends, although not significant, were observed for pro-inflammatory cytokines and chemokines Ccl3, Tnf, Cxcl1 and IL-1b (Brown et al., 2010), when comparing vehicle to boldine-treated SCI samples (Supplementary Figures 9A–D).

We then evaluated changes in the expression of Cd68 and Iba1, both makers of activated microglia (Xuan et al., 2019); Gfap, a highly expressed in reactive astrocytes (Brenner, 2014); and Mmp9 which encodes for an enzyme with degradation of extracellular matrix activity and protection of motor neuron death (Spiller et al., 2019). Levels of Cd68, Iba1, Gfap and Mmp9 were significantly increased in samples from vehicle-treated injured spinal cords in comparison to shams. However, these changes were not observed in samples from injured animals treated with boldine, and were significant when comparing vehicle-treated versus boldine-treated SCI groups (Figures 6D–G; p < 0.01 and p < 0.05). Levels of Cx43 also increased after injury (Supplementary Figure 9E; p < 0.0001), but they did not change when comparing sham to boldine-treated SCI samples (Supplementary Figure 9E; p > 0.05).

We also tested for genes that encode for SNAP25, involved in synaptogenesis (Antonucci et al., 2016), and Gap43 involved in axonal growth and plasticity (Koshi et al., 2010), and for the NMDA receptor subunit, Grin2b (Zhong et al., 1995). Interestingly, levels of Snap25 and Gap43 were significantly higher in samples from SCI animals treated with boldine as compared to vehicle-treated SCI samples (Figures 6G–I; p < 0.05). A similar trend was observed for Grin2b although not significant (Supplementary Figure 9F; p > 0.05). In addition, we evaluated changes for genes NefH and Ngf, also involved in axonal function, growth, maintenance and neuron survival (Ryden et al., 1997; Hayakawa et al., 2012; Turney et al., 2016; Fornaro et al., 2020) although changes between vehicle and boldine SCI samples were not significant and only showed a trend (Supplementary Figures 9G, H; p > 0.05).

We further performed RNAscope in-situ hybridization assay to detect mRNA transcripts and validate the changes observed for Gfap, S100a8, and Cx43 using transverse cryo-sections of spinal cords immediately caudal from the lesion epicenter of boldine and vehicle treated SCI animals at 14 dpi (Figure 7). Our results showed, as expected, significantly higher levels of Gfap and S100a8 in samples from vehicle-treated SCI animals as compared to boldine-treated SCI animals (Figures 7C, G; p < 0.05). However, levels of Cx43 transcripts remain similar at this timepoint (Figure 7D; p = 0.7812).