Spinal cord–protective effect of resorcimoline for ischemia-reperfusion injury in a rabbit model

Abstract

Objectives

Spinal cord injury can occur after aortic surgery as a result of reactive oxygen species (ROS) generated during ischemia-reperfusion. Resorcimoline is a novel and potent antioxidant that has demonstrated ROS-scavenging activity in vitro. We evaluated the spinal cord–protective effects of resorcimoline and compared its efficacy with that of edaravone, a free-radical scavenger currently used clinically in Japan.

Methods

A spinal cord injury model was established via infrarenal aortic occlusion for 20 minutes in 30 Japanese white rabbits. The animals received saline (control), resorcimoline (6 mg/kg), or edaravone (6 mg/kg) immediately after reperfusion. Neurologic function was assessed using the modified Tarlov score, and a linear mixed model was used to analyze the modified Tarlov score data. Logistic regression analysis was used to compare the risk of paraplegia between groups. After 5 days, spinal cord segments (L3-L5) were harvested. The number of residual normal neurons was evaluated histologically, and oxidative DNA damage was assessed via immunostaining for 8-hydroxy-2-deoxyguanosine.

Results

Compared with the control group, resorcimoline treatment resulted in significantly greater modified Tarlov score improvement (regression coefficient; 2.025; 95% confidence interval, 0.267-3.782; P = .025), lower incidence of paraplegia (odds ratio, 0.04; 95% confidence interval, 0.002-0.4; P = .004), greater number of residual normal neurons (P = .018), and greater percentage of 8-hydroxy-2-deoxyguanosine−negative neurons (P < .001). Conversely, these parameters did not differ significantly between the edaravone and control groups.

Conclusions

Resorcimoline demonstrated spinal cord–protective effects, likely associated with its antioxidant ability to eliminate ROS generated during ischemia-reperfusion.

Graphical Abstract

Video Abstract

Video 1

Brief illustration of the spinal cord ischemia (SCI) model creation and the experimental protocol. Video available at: https://www.jtcvs.org/article/S2666-2736(25)00386-9/fulltext.

Resorcimoline attenuates paraplegia after spinal cord ischemia-reperfusion.

Central Message.

Prophylactic administration of resorcimoline, a novel antioxidant, significantly attenuates spinal cord injury in a rabbit ischemia-reperfusion model.

Perspective.

This study provides rare in vivo evidence supporting the neuroprotective efficacy of resorcimoline, a novel antioxidant, in a rabbit spinal cord ischemia model. Given the critical role of oxidative stress in spinal cord injury after aortic surgery, these findings may contribute to the development of preventive strategies to reduce paraplegia—a severe complication with limited therapeutic options.

Most aortic aneurysm surgeries are performed to prevent rupture in patients who are asymptomatic; therefore, patients are generally expected to recover without functional impairment. However, ischemic spinal cord injury (SCI) complicates 2% to 15% of open-repair cases¹ and 11% of endovascular repair cases² for thoracoabdominal aneurysms, despite advancements in surgical techniques and preventive strategies. During open repair of an aortic aneurysm, extensive ischemia and subsequent reperfusion of the spinal cord and internal organs can lead to ischemia-reperfusion injury (IRI), which may cause more severe damage than ischemia alone. Oxidative stress, reactive oxygen species (ROS)-induced inflammation, and IRI-related cytokine release can significantly exacerbate SCI.^3,4 Despite multiple strategies, no proven therapy exists for IRI-related SCI.⁵ Edaravone, a free-radical scavenger, has been approved for clinical use in Japan but only for acute cerebral infarction and amyotrophic lateral sclerosis.^6,7 Although a single case report of edaravone in ischemic SCI has been published,⁸ no beneficial effect was observed, and no effective antioxidant therapy has yet been established for the clinical treatment of SCI. Our previous in vitro studies demonstrated that resorcimoline (RML; International Union of Pure and Applied Chemistry name: 5-[acetyl-methyl-(4-hydroxyphenyl)methyl]benzene-1,3-diol), a newly synthesized antioxidant, exhibited broad ROS-scavenging activity and was more potent than edaravone, particularly against superoxide anion and nitric oxide.⁹ In addition, an in vivo study using a lower-limb ischemia model showed that RML protected lower-extremity motor function and muscle fiber tissue from IRI.¹⁰ The present study aimed to evaluate the spinal cord–protective effects of RML in vivo using a rabbit SCI model and to compare its efficacy with that of edaravone.

Methods

Resorcimoline

RML was fortuitously synthesized while attempting to synthesize dimethylresveratrol and dimethylpiceatannola. The method for producing RML was established by TechnoPro R&D Company, which prepared and provided the samples used in this study.

SCI Models

This study was conducted in accordance with the Animal Experimentation Regulation Law and was approved by the Oita University Animal Ethics Committee (approval number: 212202). All animals were treated according to the animal experiment guidelines of Oita University. Each animal was kept in a single cage with a 12/12-hour light/dark cycle and given ad libitum access to water and food.

The SCI model used male Japanese white rabbits (n = 30). Anesthesia was introduced by administering a 50 mg/kg ketamine intramuscular injection, followed by 5% isoflurane and oxygen inhalation. The rabbits were then maintained on 2.0% to 3.0% isoflurane and oxygen inhalation. The animals breathed spontaneously via face masks and were maintained supine on a heating pad maintaining body temperature within the normal rabbit range (38 °C-39.5 °C). A venous line was secured in the auricular vein through which maintenance fluid was administered (saline solution, 4 mL/h).

A skin incision was made in the groin to expose the common femoral artery, into which a 4-Fr Ultra high-flow sheath (MEDIKIT Co) was inserted. The lower-body blood pressure was continuously monitored through the side port of the sheath. A 3-Fr Fogarty balloon catheter (Edwards Lifesciences) was inserted into the sheath and positioned into the aorta 15 cm from the inguinal ligament. Based on methods from previous literature, the distal end of the balloon catheter was positioned 0.5 to 1.5 cm distal to the left-renal artery.¹¹ The upper-body blood pressure was continuously monitored as the pressure at the catheter tip. Afterward, the aorta was occluded for 20 minutes by inflating the balloon until the lower-body blood-pressure waveform completely disappeared.

Among the 30 rabbits, 10 each were assigned to the control, RML, and edaravone groups. After the 20-minute occlusion period, the balloon was deflated to facilitate reperfusion. Immediately after reperfusion, the respective treatments of each group were administered intravenously for 3 minutes. The control group was given 12 mL of saline solution, and the RML group was given a solution of 6 mg/kg RML dissolved in 2 mL of distilled water, to which 10 mL of saline was added for a total volume of 12 mL. The edaravone group was given 12 mL of 6 mg/kg edaravone (Tanabe Mitsubishi Pharma). Blood pressure and rectal temperature were recorded every 5 minutes from balloon inflation until 20 min post deflation. During balloon inflation, both upper- and lower-blood pressure were recorded.

After completing these procedures, all catheters were removed, and the animals were returned to their cages after waking from anesthesia. Hind-limb neurologic function was evaluated daily on days 0 to 5. Animals were maintained for 5 days in accordance with a specifically approved protocol, with supportive care provided throughout. On day 5, the animals were euthanized via rapid intravenous injection of potassium under deep anesthesia. Immediately after the euthanasia, the L3–L5 vertebrae were removed, and the spinal cord was harvested and fixed in 10% formalin (Video 1).

Neurologic Evaluation

The modified Tarlov Score (MTS)¹² was used to evaluate hind-limb neurologic function, categorized as follows: 0, no movement; 1, slight movement; 2, sit with support; 3, sit alone; 4, weak hop; and 5, normal hop. An MTS of 0 was defined as complete paraplegia. The animals were evaluated and categorized daily on days 0 to 5, and the change in MTS was analyzed as continuous numerical data. The incidence of complete paraplegia during the entire observation period was determined.

Count of Residual Normal Neurons

The L3 to L5 spinal cords were paraffin-embedded and subjected to hematoxylin-eosin and Klüver-Barrera staining. One section each from L3, L4, and L5 was prepared, and the entire spinal cord area was imaged at 10 × magnification and reconstructed to allow counting of residual normal neurons across the whole cross-section. Residual normal neurons were defined as neurons with round basophilic nuclei and distinct Nissl bodies in the gray matter. Three researchers, blinded to the study groups and neurologic outcomes, were tasked to count the average total number of residual normal neurons in the L3 to L5 regions for analysis.

8-OHdG Immunostaining

A commonly used marker of oxidative damage is 8-hydroxy-2′-deoxyguanosine (8-OHdG), an oxidatively damaged base that is produced when guanine is oxidatively modified.^13,14 The L3 to L5 spinal cord sections were paraffin-embedded and subjected to 8-OHdG immunostaining using a mouse monoclonal antibody against 8-OHdG (Anti-DNA/RNA Damage Antibody [15A3]; StressMarq Biosciences Inc). Neurons with brown-stained nuclei were 8-OHdG-positive, whereas those with blue-stained nuclei (without brown staining) were 8-OHdG-negative. The number of 8-OHdG-positive and -negative neurons in the L3 to L5 regions, as well as the percentage of 8-OHdG-negative neurons, were evaluated.

Statistical Analysis

The physiologic variables were described as the mean ± standard deviation followed by analysis using the Steel test. The differences between each experimental group versus the control group were compared. A linear mixed model was used to analyze the MTS values in each group, in which the treatment group was included as a fixed effect, and the day of measurement was treated as a random effect. The Dunnett multiple comparison test was used to test the differences between the experimental and control groups. Binary logistic regression was performed to evaluate the association between each group and the risk of developing paraplegia throughout the entire observation period. We performed a receiver operating characteristic analysis to evaluate the discriminative ability of the logistic regression model using RML and edaravone as predictors for paraplegia occurrence (paraplegia = 1). The area under the curve was calculated and the 95% confidence interval (CI) was estimated using bootstrap resampling. The number of residual normal neurons and the percentage of 8-OHdG-negative neurons were described as the median and interquartile range (IQR) followed by performing the Steel test for multiple comparisons to compare the differences between the experimental groups versus the control group. JMP Pro 18 (SAS Institute Inc) was used for all statistical analyses.

Results

Physiologic Variables

The physiologic variables are presented in Table 1. There were no significant differences between either experimental group and the control group in terms of body weight, intraoperative systolic blood pressure, or rectal temperature.

Table 1.

Physiologic variables

Variable	Control	RML	Edaravone	P (control vs RML/control vs edaravone)
Body weight, kg	2.8 ± 0.1	2.9 ± 0.1	2.9 ± 0.3	.584
				.974
Rectal temperature, °C
Before occlusion	39.3 ± 0.5	39.3 ± 0.8	39.3 ± 0.4	.901
				1.000
During occlusion	38.6 ± 0.5	39.0 ± 0.8	38.6 ± 0.5	.087
				.665
After reperfusion	38.3 ± 0.4	38.8 ± 0.8	38.3 ± 0.3	.104
				.996
Blood pressure, mm Hg
Before occlusion	98.5 ± 23.1	96.4 ± 15.5	92.6 ± 38.9	.991
				.974
During occlusion (proximal)	99.2 ± 17.5	98.5 ± 13.7	100.6 ± 17.5	.810
				.991
During occlusion (distal)	25.8 ± 17.3	22.0 ± 8.4	20.1 ± 5.8	.919
				.910
After reperfusion	93.9 ± 23.5	88.6 ± 4.4	93.4 ± 20.1	.679
				.997

Variables are presented as the mean ± standard deviation. Comparisons between groups were performed using the Steel test. RML, Resorcimoline.

Neurologic Evaluation

The number of rabbits in each MTS category per day is shown in Figure 1, A. Compared with the control group, the change in MTS from days 0 to 5 was significantly greater in the RML group (regression coefficient: 2.03; 95% CI, 0.27-3.782; P = .026), whereas no significant difference was observed in the edaravone group (regression coefficient: 0.40; 95% CI, −1.36 to 2.16; P = .644) (Table 2). The incidences of complete paraplegia in the control, RML, and edaravone groups were 70% (7/10), 10% (1/10), and 60% (6/10), respectively. The odds ratio for complete paraplegia compared with the control group was 0.04 (95% CI, 0.002-0.4; P = .004) for the RML group and 0.6 (95% CI, 0.09–4.1; P = .638) for the edaravone group (Table 3). The receiver operating characteristic curve is shown in Figure 1, B, with an area under the curve of 0.723 (95% CI, 0.535-0.888). In both analyses, the RML group showed a significant difference compared with the control group, whereas the edaravone group did not.

Figure 1.

Neurologic evaluation. A, Daily number of animals in each MTS category The number of animals in each modified Tarlov score (MTS) category is shown per day from days 0 to 5. The MTS categories are as follows: 0, no movement; 1, slight movement; 2, sit with support; 3, sit alone; 4, weak hop; and 5, normal hop. An MTS of 0 indicates complete paraplegia. B, Receiver operating characteristic (ROC) curve The ROC curve of logistic regression analysis of the incidence of complete paraplegia is shown, and the area under the ROC curve is 0.723 (95% confidence interval, 0.535-0.888).

Table 2.

Differences in MTS transition from day 0 to day 5 analyzed using a linear mixed model

Group	Regression coefficient	95% confidence interval	Standard error	P value
Control vs RML	2.03	0.27 to 3.78	0.86	.026
Control vs edaravone	0.40	−1.36 to 2.16	0.086	.664

Differences in the transition of MTS values in each group from day 0 to day 5 were analyzed using a linear mixed model, and differences between the experimental and control groups were tested using the Dunnett multiple comparison test. Bold numbers indicate P values < .05. RML, Resorcimoline.

Table 3.

Incidence of complete paraplegia throughout the entire observation period

Incidence of complete paraplegia
Control	70% (7/10)
RML	10% (1/10)
Edaravone	60% (6/10)

Logistic regression analysis of the incidence of complete paraplegia	Odds ratio	95% CI	P value
Control vs RML	0.04	0.002-0.4	<.01
Control vs edaravone	0.64	0.09-4.1	.638

Bold numbers indicate P values < .05. RML, Resorcimoline; CI, confidence interval.

Count of Residual Normal Neurons

The typical findings of hematoxylin–eosin and Klüver-Barrera staining in each group are shown in Figure 2, A. The median (IQR) numbers of residual normal neurons were 24.1 (2.1-30.3), 30.5 (24.5-36.4), and 20.5 (0.5-25.2) in the control, RML, and edaravone groups, respectively, as shown in Figure 2, B. Compared with the control group, the median (IQR) number of residual normal neurons was significantly greater in the RML group (P = .018) but was not significantly different in the edaravone group (P = .536).

Figure 2.

A, Histologic findings. The typical findings of hematoxylin–eosin (upper rows) and Klüver–Barrera staining (lower rows) in each group are shown (scale bar: 100 μm). Control group: The neurons disappeared, whereas glial cells, foam cells, and lymphocytes gathered. RML group: Neuronal nuclei and Nissl bodies were visible, with clear cytoplasmic basophilic staining. Edaravone group: Neurons were present, but some had no nuclei. The Nissl bodies disappeared, and the cytoplasmic basophilic staining appeared to have faded. B, Number of residual normal neurons. The numbers of residual normal neurons are presented. The lower and upper borders of the box represent the lower and upper quartiles (25th percentile and 75th percentile, respectively). The middle horizontal line represents the median. The lower and upper whiskers represent the minimum and maximum values, respectively. The Steel test for multiple comparisons was used to compare the differences between the experimental and control groups. Bold numbers indicate P values < .05.

8-OHdG Immunostaining

Representative 8-OHdG immunostaining findings for each group are shown in Figure 3, A. The median (IQR) percentages of 8-OHdG–negative neurons were 0.0% (0.0%-5.4%), 10.0% (4.1%-16.2%), and 0.0% (0.0%-7.5%) in the control, RML, and edaravone groups, respectively, as shown in Figure 2, B. Compared with the control group, the RML group had a significantly greater median percentage of 8-OHdG–negative neurons (P < .001), whereas the difference in the edaravone group was not statistically significant (P = .672).

Figure 3.

8-OHdG immunostaining findings The typical 8-OHdG immunostaining findings of each group are shown (scale bar: 100 μm). A, 8-OHdG−positive neurons in the control and edaravone groups are shown. Brown-stained nuclei (arrow) were considered to show positive 8-OHdG expression. B, 8-OHdG-negative neurons in the resorcimoline group. Blue-stained nuclei (triangle) were considered to show negative 8-OHdG expression. C, Percentage of 8-OHdG-negative neurons The percentages of 8-OHdG-negative neurons are presented. The lower and upper borders of the box represent the lower and upper quartiles (25th percentile and 75th percentile, respectively). The middle horizontal line represents the median. The lower and upper whiskers represent the minimum and maximum values, respectively. The Steel test for multiple comparisons was used to compare the differences between the experimental and control groups. Bold numbers indicate P values < .05.8-OHdG, 8-Hydroxy-2′-deoxyguanosine.

Discussion

The rabbit aortic balloon occlusion model has been widely used and validated in previous SCI studies. Earlier work has tested a range of ischemic durations from 10 to 30 minutes. Shorter intervals (10-15 minutes) often produce incomplete injury or high variability, whereas longer intervals (30 minutes or more) frequently result in irreversible damage.¹¹ We selected a 20-minute occlusion in our study to balance severity with reproducibility. The dose of 6 mg/kg for both RML and edaravone was selected on the basis of previous reports of edaravone use in rabbit SCI models.15, 16, 17

In this exploratory study, RML markedly attenuated the severity of spinal cord ischemia–reperfusion injury compared with controls. Neurologic function was preserved, paraplegia was avoided, and significantly more residual normal neurons were observed in the anterior horn of the RML group. In contrast, edaravone did not provide measurable benefit under the present experimental conditions.

From a mechanistic standpoint, edaravone primarily scavenges hydroxyl radicals generated downstream,⁶ whereas RML efficiently captures upstream radicals such as superoxide and nitric oxide,⁹ thereby preventing the formation of highly toxic species including peroxynitrite and hydroxyl radicals.¹⁸ Blocking oxidative injury at an earlier stage likely contributed to the superior effect of RML in this stringent model. Consistent with this, 8-OHdG analysis demonstrated that more neurons were spared from oxidative DNA damage in the RML group. Nevertheless, it should be noted that the proportion of 8-OHdG–negative neurons remained relatively modest compared with the striking preservation of hindlimb motor function. This discrepancy suggests that protection cannot be explained by attenuation of oxidative DNA damage alone. Other mechanisms may have contributed, including survival of key neuronal populations, partial circuit reorganization, or neuroplastic processes such as synaptic remodeling, dendritic spine formation, and axonal sprouting.¹⁹ Furthermore, the lack of efficacy of edaravone, despite its established role as a radical scavenger, raises the possibility that RML protects the spinal cord through additional mechanisms that remain to be clarified. Whether RML primarily reduced ischemic damage itself or suppressed reperfusion injury is also uncertain. Mechanistic analyses such as inflammatory cytokine profiling, apoptosis, and autophagy/mitophagy assays were not performed due to tissue limitations. Incorporating these approaches in future studies may help clarify the mechanisms of action of RML.

Several issues remain to be addressed before RML can be optimally applied. Dosing and timing are critical. The regimen described in this study, although effective, may not be optimal. Alternative strategies, such as elevating plasma levels before clamp release,¹⁶ repeated or continuous administration during early reperfusion,¹⁷ or delayed dosing to coincide with oxidative stress peaks,¹⁶ may also prove beneficial. Studies are needed to define the optimal regimen. Our rabbit model replicates the essential ischemia-reperfusion insult seen in aortic repair but does not incorporate important clinical modifiers. These include collateral circulation, interruption of segmental arteries, comorbidities, and adjunctive measures such as cerebrospinal fluid (CSF) drainage, staged repair, or distal perfusion, as well as hemodynamic variations during cross-clamping. While the model isolates reperfusion biology under controlled conditions. Clinically, CSF drainage and proximal pressure management remain indispensable,⁵ but antioxidant therapy such as RML could potentially complement these approaches.

Safety must also be considered. Edaravone is associated with adverse events including hepatic and renal dysfunction, hypersensitivity.²⁰ RML has no history of clinical use and very limited toxicologic or pharmacokinetic data. At least no adverse effects were observed in this study.

Finally, previous studies have shown protective effects of edaravone in rabbit spinal cord ischemia models, mainly with shorter ischemia or repeated/continuous dosing.15, 16, 17 In contrast, our study used a single bolus immediately after reperfusion under more severe conditions, where edaravone did not show a significant effect and its efficacy may have been underestimated. Given the different mechanisms of action,⁹ combining RML and edaravone may yield additive or synergistic effects. Future studies should test validated edaravone regimens and combination strategies for fairer comparison.

Limitations

Several limitations should be acknowledged. First, this was an exploratory study with relatively small group sizes (n = 10 per arm), which were set a priori to balance the 3R principle (reduction) with feasibility and to align with precedent in rabbit spinal cord ischemia models. No prospective power analysis was performed; variance estimates from the present dataset will inform sample-size calculations in future confirmatory studies. Second, mechanistic assays were not conducted, limiting insights into the pathways through which RML exerts its protective effects. Third, the optimal dose/timing remain undefined, and pharmacokinetics and long-term safety are lacking. Fourth, translation to humans is limited by species differences: the rabbit spinal cord depends largely on homosegmental blood supply, making it especially vulnerable to ischemia.²¹ Interspecies vascular differences, patient heterogeneity, surgical variability, and pharmacologic factors have hindered the clinical translation of many otherwise-promising neuroprotective strategies.

Conclusions

RML demonstrated potent spinal cord–protective effects, which may be attributed to its strong antioxidant properties, specifically its ability to eliminate ROS generated during ischemia-reperfusion. These protective effects were observed even under a severe ischemia-reperfusion model, with significant neurologic improvement despite modest histologic differences, highlighting the importance of both neuronal preservation and functional network protection. RML thus represents a promising candidate for adjunctive therapy alongside established clinical strategies such as CSF drainage and perfusion management. These findings may contribute to the development of preventive strategies against paraplegia.

Conflict of Interest Statement

A patent application related to RML by Drs Tokumaru, Higuchi, and Miyamoto is currently pending in Japan (no. 2023-119682). The authors and Oita University have no financial interest in this compound, and no commercial development is underway at present. All other authors reported no conflicts of interest.

The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.

Supplementary Data

Video Abstract

A Video Abstract is available to provide a brief overview of the objectives, methodology, key results, and implications of this study.