Mitigation of Acute Seizures and Neuropathology after Traumatic Brain Injury by Structure-Based Discovery-Identified Drugs

Abstract

Our objective was to test the hypothesis that structure-based identified or designed compounds exhibiting neuroprotective, antioxidant, and anti-inflammatory properties in vitro will mitigate early seizures and neuropathology after traumatic brain injury (TBI) in vivo. The neuroprotective and anti-inflammatory effects of 11 compounds identified by computer-assisted approximations were tested in vitro in neuronal microglial co-cultures. Among these, compound FBA exhibited the best neuroprotective (MAP-2, microtubule-associated-protein 2, a neuronal damage biomarker), antioxidative (nitrite production), and anti-inflammatory effects in vitro (all p < 0.01). Consequently, its neuroprotective and antiseizure effects were assessed in vivo in adult male Sprague-Dawley rats exposed to severe lateral fluid-percussion-induced TBI. Rats under continuous video-electroencephalogram monitoring received prophylactic treatment with an intraperitoneal (i.p.) injection of FBA (FBApro, 30 mg/kg) or vehicle (VEH, 48% PEG in 0.9% saline, 3 mL/kg) at 2 and 24 h post-TBI. Rats that developed status epilepticus received 1–2 additional on-demand FBA doses (FBApro+, 100 mg/kg, i.p.) to stop epileptiform activity. FBApro treatment reduced the cortical lesion area (18.9 ± 4.1 mm², n = 7) compared with VEH treatment (24.8 ± 5.7 mm², n = 10, p < 0.05). FBApro treatment also showed a favorable effect on the white matter by reducing plasma levels of pNF-H, a biomarker of axonal injury, compared with VEH treatment (Cohen’s delta 0.657). Both FBApro (368 ± 407 s) and FBApro+ (256 ± 327 s) treatments reduced the average cumulative seizure duration compared with VEH (896 ± 703 s, both p < 0.05). The FBApro+ treatment regimen also reduced the mean relative theta and alpha power and increased the mean relative gamma power in the electroencephalogram (p < 0.05). Our data identified FBA as a novel structure-based discovered compound with promising favorable effects on structural and functional recovery after TBI.

Introduction

Annually, ∼69 million individuals suffer from traumatic brain injury (TBI) worldwide, and TBI is a prominent cause of injury-related death globally.^1,2 As the primary injury cannot be reversed, therapeutic interventions target the secondary brain damage that occurs within minutes after the primary impact, including neurotoxicity, oxidative stress-related alterations, and neuroinflammation, which progress in a parallel and/or sequential manner.^3
–5 Importantly, the secondary damage occurs over hours to days to weeks postinjury, providing a window of opportunity for therapeutic interventions to prevent the development of functional impairments such as acute seizures or post-traumatic epilepsy (PTE).^6
–8

Up to 30% of adults with severe TBI develop acute seizures.^9

–13 Our recent pre-clinical study showed that 70–80% of rats with severe TBI induced by lateral fluid-percussion (LFP) injury, a model of human closed head injury, developed seizures and status epilepticus (SE) during the first-week postinjury.¹⁴ Although the long-term consequences of post-TBI acute seizures and SE are incompletely understood, previous epidemiological and observational studies suggest that acute post-TBI seizures are associated with a greater probability of mortality and development of PTE.^8,12,15,16

Despite a large number of hypothesis-driven drug-discovery studies, the treatment of TBI remains a major unmet medical need.^17
–19 Computer-aided drug discovery, including virtual screening and structure-based drug design, involves the use of information technologies to identify and/or modify, on a rational basis, active scaffolds with the desired bioactivity profiles.²⁰ Computer-aided drug discovery has been successfully applied to identify treatments for hypertension Parkinson’s disease, neuropathic pain, and HIV (Human immunodeficiency virus),^21
–23 among other conditions. Currently, more than 20 artificial intelligence-derived drug candidates are undergoing clinical trials related to a diversity of disorders, such as cancer, Alzheimer’s disease, and diabetes.²⁴

Here, we screened a series of computationally identified repurposed drugs and drug candidates to investigate their potential to mitigate TBI-induced brain damage, acute seizures, and epileptiform activity. We hypothesized that compounds with an ability to reduce neuronal death, oxidative stress, and neuroinflammation in vitro could alleviate acute post-TBI pathology and promote post-TBI structural and functional recovery in vivo. The compounds identified using computational approaches were acquired or synthesized and tested in vitro in neuronal microglial co-cultures. The most promising compound, FBA, was then investigated in vivo in the LFP injury model to assess whether it would suppress the occurrence of acute postimpact electrographic seizures, reduce the extent of cortical lesion area, and lower the plasma levels of axonal injury marker, pNF-H. Our proof-of-concept data demonstrate that structure-based compound discovery is a promising approach to discovering novel treatments for TBI.

Materials and Methods

The study designs are summarized in Figures 1 (in vitro) and 2 (in vivo). Table 1 summarizes the compounds investigated in this study and their presumed mechanisms of action and antiseizure effects. Figure 3 displays the randomization of rats into different treatment groups, reasons for exclusions, and study flow. Figure 4 shows the electrode montage used. Detailed descriptions of the research methods used to perform the in vitro and in vivo experiments, including ethics and statistics are provided in Supplementary Data.

FIG. 1.

In vitro study design. On day 0, E18 mouse brains were dissected, and cortical neuronal cells were plated. On day 5, BV2 microglial cells were added on top of the neurons. After 1 h, the co-cultures were pretreated with the test compounds for 60 min. Then, LPS and IFNγ were added to the cell culture medium to induce neuroinflammation. On day 7 (i.e., after a 48-h exposure to neuroinflammation), the cell culture medium was collected for analysis of secreted TNFα and nitrite levels in the medium. Co-culture cells were fixed for analysis of neuronal viability using a MAP-2 immunoassay. IFNγ, interferon gamma; LPS, lipopolysaccharides; MAP-2, microtubule-associated-protein 2; TNFα, tumor necrosis factor alpha.

FIG. 2.

In vivo study design. The neuroprotective and antiseizure effects of FBA were investigated in a rat TBI model induced by LFP injury. EEG-recording electrodes were implanted during the same surgery. Recording of vEEG was started immediately and continued for 14 days. (A) Cohort 1—study design feasibility, neuroprotection, and antiseizure efficacy (preliminary study). Vehicle or FBA (30 mg/kg) was administered prophylactically at 2 and 24 h after TBI to determine the safety and antiseizure efficacy of the FBA. Tail vein blood was collected at 72 h post-TBI to assess the effect of FBA on plasma levels of phosphorylated neurofilament heavy chain. (B) Cohort 2A—antiseizure efficacy. Vehicle or FBA (30 mg/kg) was administered prophylactically at 2 and 24 h after TBI to increase the cohort size to assess the efficacy of FBA on acute seizures. (C) Cohort 2B—additional dosing to suppress status epilepticus (SE). As we noticed that some animals developed SE despite prophylactic treatment, we supplemented the prophylactic treatment with 1–2 additional doses of FBA (or vehicle). The vEEG was monitored online, and if epileptiform activity fulfilling the Salzburg criterion of SE was detected,²⁵ rats were treated with 100 mg/kg (i.p.) of FBA or vehicle. If SE did not stop or re-occurred later, the animal was treated with one additional dose of FBA (30 mg/kg, i.p.) or vehicle, no earlier than 30 min after the first additional dose. At the end of the experiment, brains were either perfused or immersion-fixed for histology. D, day; dH₂O, distilled water; DMSO, dimethyl sulfoxide; PBS, phosphate buffered saline; PEG, polyethylene glycol; SE, status epilepticus; TBI, traumatic brain injury; vEEG, video-electroencephalogram.

Table 1.

Compounds Investigated in the Present Study, Their Abbreviations, Presumed Mechanisms of Action, Antiseizure Efficacy, and Adverse Events

Compound	Compound name	MoA	In vitro effect	In vivo antiseizure test—efficacy	Rotarod	Reference
BP1	Benzylparaben	Nav1.2 blocker Cosmetic preservative Antifungal properties	Nav1.2 blocker, in a concentration-dependent manner			Maier et al. 2009²⁶ Juliano and Magrini 2017²⁷ Enrique et al. 2020²⁸
CA1	Sodium cyclamate	Possible activity on mGluR	Mild toxicity to colon cancer cell line at 100 µM	MES (+, mild)Increased latency to seizure in zebrafish	No toxicity	Talevi et al. 2012²⁹ van Eyk. 2014³⁰ Pisera-Fuster et al,2017³¹
DP1	Donepezil	Acetylcholinesterase inhibitorPredicted as GABA transporter 1 inhibitor	Neuroprotective to cortical and septal neurons	MES (+)6 Hz and PTZ mouse model (+)	No toxicity	Akasofu et al. 2008³² Kimura et al. 2012³³ Couyoupetrou et al. 2014³⁴
FBA	N-(p-fluoro) benzyl-2-hydroxyisobutylamide	Intermediate w/o Nav1.2 blocking effect	No effect on Nav1.2	MES (+)	No toxicity	Sabatier et al. 2019³⁵
FBS	3-p-fluorbenzyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-one-2,2-dioxide	Nav1.2 blocker	Nav1.2 blocker in HEK cells	MES (+)	No toxicity	Sabatier et al. 2019³⁵
MLV7	3-[(2-propylpentyl) sulfamoylamino] propanoic acid	Carbonic anhydrase VII inhibitor				Unpublished
MLV12	Methyl [N-(N’-4-fluorobenzyl)]-sulfamoyl] l-valinate	Carbonic anhydrase VII inhibitor	Medium affinity to benzodiazepine-binding site (GABA_A receptor) in synaptosomal membranes	MES (+)	Not performed	Villalba et al. 2016³⁶ Villalba et al. 2016³⁷
MTK	Montelukast	TRPV1 inhibitorAnti-inflammatory profileLeukotriene receptor antagonistAntioxidant		PTZ rat model (+)MES (+, mild)	No toxicity	Bousquet et al. 2009³⁸ Al Saadi et al. 2011³⁹ Cevik et al. 2015⁴⁰ Llanos et al. 2022⁴¹ Unpublished
NVB	Novobiocin	TRPV1 inhibitorAntibiotic		MES (+)	No toxicity	Rodríguez-Cerrato et al. 2010⁴² Llanos et al. 2022⁴¹ Unpublished
SA1	Sebacic acid	Selected in a phenotypic virtual screen, pursuing antiseizure, and anti-inflammatory agents		MES (+, mild)PTZ rat model (+)	No toxicity	Enrique et al. 2021⁴³
SCH	Proglumide	Cholecystokinine antagonistGABA transaminase inhibitor		Rat Lithium-Pilocarpine induced SE model (+)	No toxicity	Ahmad et al, 2014⁴⁴ Ahmad and Wadaan 2014⁴⁵ Unpublished

+, reduced seizure activity. Data were obtained from references cited in the rightmost column.

GABA, gamma-aminobutyric acid; mGluR, metabotropic glutamate receptor; Nav1.2, sodium channel 1.2 isoform; PTZ, pentylenetetrazol; SE, status epilepticus; TRPV1, transient receptor potential vanilloid.

FIG. 3.

Randomization and study flow. A total of 54 rats experienced lateral fluid-percussion-induced traumatic brain injury (TBI). Of the 54 rats, 22 were included in Cohort 1 (preliminary study). After acute exclusion of 3 rats, the remaining 19 animals were randomized to either prophylactic treatment with vehicle (TBI-VEHpro, n = 10) or FBA (TBI-FBApro, n = 9). All 19 rats completed the day (D) 0-D7 video-electroencephalogram (vEEG) monitoring. Of the 54 rats, 32 were included in Cohort 2. After acute exclusions of 4 animals, the remaining 28 rats were divided into Cohorts 2A and 2B. Animals in Cohort 2A (n = 15) were randomized (as in Cohort 1) to vehicle (TBI-VEHpro, n = 4) or FBA (TBI-FBApro, n = 11) treatments to increase the statistical power of data obtained in Cohort 1. All 15 rats in Cohort 2A completed the D0-D7 vEEG monitoring. Rats in Cohort 2B (n = 13) received the prophylactic treatment (as Cohorts 1 and 2A) supplemented with an on-demand administration of vehicle (TBI-VEHpro+, n = 3) or (TBI-FBApro+, n = 10) to stop status epilepticus (or ictal-interictal continuum) when diagnosed during the online vEEG. Rats were allocated to the TBI-VEHpro+ or TBI-FBApro+ group if the online EEG analysis indicated an ongoing status epilepticus (>5 min). After 2 exclusions, 11 rats in Cohort 2B completed the D0-D7 vEEG monitoring. Reasons for exclusions at each follow-up stage are shown in gray shading (number of animals is in parenthesis).

FIG. 4.

Electrode montage. (A) Epidural cortical screw electrodes (C3, C4, O1, and O2) were implanted in 2/19 rats in Cohort 1, in 28/28 in Cohort 2 (B) The remaining 17/19 rats in Cohort 1 were implanted with both cortical screw and intracerebral depth (A1, A2, Po1, Po2, H1, and H2) electrodes. Two epidural screw electrodes served as ground (G) and reference (Ref) electrodes. The orange dot indicates the center of the craniectomy (orange circle). Grid dimensions: 1 mm × 1 mm.

Results

In vitro validation of the effects of the compounds on neuronal viability, oxidative stress, and neuroinflammation

Neuronal viability

To determine the effects of the compounds on neuronal survival, a neuronal viability immunoassay was performed. The percentage of neuronal viability in neuronal-BV2 microglial co-cultures exposed to neuroinflammatory mediators LPS and interferon gamma (IFNγ) only (LPS/IFNγ+) was set to 0%. In co-cultures treated also with positive control (iNOS inhibitor 1400W), the percentage of neuronal viability was set to 100%. The neuroprotective effect of each test compound was compared with that in co-cultures treated with LPS/IFNγ+ only. Data are summarized in Figure 5A, Supplementary Figures S1A, B and S2.

FIG. 5.

Validation of the neuroprotective, antioxidant, and anti-inflammatory effects of compounds discovered through structure-based design. (A) Neuronal viability. To facilitate comparative analysis of neuronal survival between the batches and compounds, the percentage of neuronal viability in cortical-BV2 microglial co-cultures exposed to LPS/IFNγ+ only was set to 0% and that in co-cultures treated with the iNOS inhibitor 1400W (positive treatment control) was set to 100%. None of the benzylparaben (BP1) concentrations improved neuronal viability (all p > 0.05 compared to LPS/IFNγ+ wells). In fact, 100 µM of BP1 was toxic to both neuronal and BV2-microglial cells (Supplementary Fig. S2F). At 100 µM, donepezil hydrochloride (DP1) improved neuronal survival from 0% to 90% (p < 0.001), at 30 µM from 0% to 72% (p < 0.001), at 10 µM from 0% to 51% (p < 0.001), and at 1 µM from 0% to 38% (p < 0.01). In microscope images, however, the wells treated with 100 µM (and even 30 µM) of DP1 had some degenerating neurons whereas the wells treated with 10 µM of DP1 had an overall healthier profile (Supplementary Fig. S2G). At 100 µM, FBA improved neuronal survival from 0% to 51% (p < 0.001). At 100 µM, 3-4-fluorbenzyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-one2,2-dioxide (FBS)increased neuronal survival from 0% to 79% (p < 0.001) and at 30 µM to 29% (p < 0.05). At 100 µM, sebacic acid (SA1) increased the neuronal viability from 0% to 33% (p < 0.05). (B) Nitrite levels. For a comparative analysis, the percentage of nitrite levels in the cortical-BV2 microglial co-cultures exposed to LPS/IFNγ+ only was set to 100% and that in co-cultures treated with the iNOS inhibitor 1400W (positive treatment control) to 0%. At 100 µM, BP1 reduced nitrite levels from 100% to 28% (p < 0.001 compared to LPS/IFNγ+ wells). This reduction of the percentage of nitrite by BP1, however, was associated with remarkable degeneration of both neurons (panel A) and microglial cells (Supplementary Fig. S2F). At 1 µM, BP1 increased nitrite levels from 100% to 116% (p < 0.05). At 100 µM, DP1 reduced the levels of nitrite from 100% to 3% (p < 0.001), which were associated with cell degeneration (Supplementary Fig. S2G). At 10 µM, DP1 increased nitrite levels from 100% to 117% (p < 0.05) and at 1 µM from 100% to 128% (p < 0.001). At 100 µM, FBA reduced the levels of nitrite from 100% to 47% (p < 0.001 compared with LPS/IFNγ+ wells). At 100 µM, FBS reduced the levels of nitrite from 100% to 38% (p < 0.001). FBS at 1 µM, however, increased nitrite levels from 100% to 116% (p < 0.05). At 100 µM, SA1 reduced the nitrite levels from 100% to 64% (p < 0.001). (C) Levels of TNFα. To facilitate comparative analysis of TNFα levels, the percentage of TNFα in the cortical-BV2 microglial co-culture wells exposed to LPS/IFNγ only (LPS/IFNγ+) was set to 100% and that in co-cultures treated with the anti-inflammatory cytokine IL10 (positive treatment control) to 0%. At 100 µM, BP1 reduced TNFα levels from 100% to −125% (p < 0.001 compared with LPS/IFNγ+ wells), but this reduction was associated with cell degeneration (panel A, Supplementary Fig. S2F). At 100 µM, DP1 reduced the levels of TNFα from 100% to −158% (p < 0.001), but this reduction was also associated with cell degeneration observed in microscope images (Supplementary Fig. S2G). At 30 µM, DP1 reduced TNFα levels from 100% to −22% (p < 0.01). At 100 µM, FBA reduced the levels of TNFα from 100% to −23% (p < 0.01). At 100 µM, FBS reduced TNFα levels from 100% to −158% (p < 0.001) and at 30 µM from 100% to 9% (p < 0.05). At 100 µM, SA1 reduced TNFα levels from 100% to −1% (p < 0.05). Statistical significance: ***p < 0.001; **, p < 0.01; *, p < 0.05 compared with LPS/IFNγ+ (linear regression in R). The number of independent experiments in panel A was 3 and in panels B and C, 4. Data are presented as box plots with whiskers (minimum and maximum; box: interquartile range; line: median). Each dot shows an experimental value. BV2−, cortical neuronal cultures without BV2-microglial cells; IL10+, co-cultures treated with the positive control anti-inflammatory cytokine interleukin 10; LPS/IFNγ−, co-culture wells without lipopolysaccharides and interferon γ-induced neuroinflammation; LPS/IFNγ+, co-culture wells with lipopolysaccharides and interferon γ-induced neuroinflammation (untreated controls); TNFα, tumor necrosis factor α; 1400W+, co-cultures treated with the positive control, inducible nitric oxide synthase (iNOS) inhibitor 1400W hydrochloride.

Controls

Neuronal survival in wells with neuronal cells only (BV2−) was greater than in LPS/IFNγ+ co-cultures (198% vs. 0%, p < 0.001). Also, neuronal survival in co-cultures that were not exposed to neuroinflammation (LPS/IFNγ−) was greater than in LPS/IFNγ+ exposed co-cultures (52% vs. 0%, p < 0.001). The anti-inflammatory cytokine IL10 did not improve neuronal survival (9% vs. 0%, p > 0.05 compared with LPS/IFNγ+ cultures).

Test compounds

Of the 11 test compounds, 4 (DP1 > FBS > FBA > SA1) showed neuroprotective effects at one or more of the concentrations tested.

At the 100-µM concentration, DP1 increased neuronal survival from 0% to 90% (p < 0.001 compared with LPS/IFNγ+). Neuronal viability was comparable to that of the positive control 1400W (90% vs. 100%, p > 0.05). Neuronal survival was higher than in wells not treated with LPS/IFNγ (LPS/IFNγ−; 90% vs. 52%, p < 0.01). Analysis of microscope images of cultured cells, however, revealed the presence of some degenerating neurons among the surviving neurons (Supplementary Fig. S2G). Lower concentrations of DP1 also promoted neuronal survival (30 µM, 72%, p < 0.001; 10 µM, 51%, p < 0.001; 1 µM, 38%, p < 0.01 compared with LPS/IFNγ+).

At the 100-µM concentration, FBS increased neuronal survival from 0% to 79% (p < 0.001 compared with LPS/IFNγ+). Neuronal survival was comparable to that of positive control 1400W (79% vs. 100%, p > 0.05). Neuronal survival was higher than in wells not treated with LPS/IFNγ (LPS/IFNγ−; 79% vs. 52%, p < 0.05). Also, at the 30-µM concentration, FBS slightly improved neuronal survival (29%, p < 0.05 compared with LPS/IFNγ+).

At the 100-µM concentration, FBA increased neuronal survival from 0% to 51% (p < 0.001 compared with LPS/IFNγ+). Neuronal survival, however, was lower than in wells treated with positive control 1400W (51% vs. 100%, p < 0.001) but comparable to that in wells not treated with LPS/IFNγ (LPS/IFNγ−; 51% vs. 52%, p > 0.05). Lower concentrations of FBA had no favorable effects on neuronal viability (p > 0.05).

At the 100-µM concentration, SA1 improved neuronal survival from 0% to 33% (p < 0.05 compared with LPS/IFNγ+). The neuroprotective effect of SA1, however, was lower than that of positive control 1400 W (33% vs. 100%, p < 0.001). Neuronal survival was comparable to that in wells not treated with LPS/IFNγ (LPS/IFNγ−; 33% vs. 52%, p > 0.05). Lower concentrations of SA1 had no favorable effects on neuronal viability (p > 0.05)

None of the BP1 concentrations investigated promoted neuronal survival (p > 0.05 compared with LPS/IFNγ+). Also, none of the concentrations of 3-((N-(2-propylpentyl)sulfamoyl)amino)propanoic acid (MLV7), MLV12, sodium cyclamate (CA1), montelukast (MTK), novobiocin (NVB), and SCH had any effect on neuronal survival (p > 0.05, Supplementary Fig. S1A, B).

Nitrite levels

The levels of nitrite released into the cell culture medium were investigated to assess the antioxidant capacity of the test compounds. The percentage of nitrite in neuronal-BV2 microglial co-cultures exposed to neuroinflammatory mediators LPS and IFNγ only (LPS/IFNγ+) was set to 100%. In co-cultures treated with the positive control (iNOS inhibitor 1400W), the percentage of nitrite was set to 0%. The antioxidant effect of each test compound was compared with that in co-cultures treated with LPS/IFNγ+ only. Data are summarized in Figure 5B and Supplementary Figure S1C, D.

Control wells

Nitrite levels in wells with neuronal cells only (BV2−) were lower than those in LPS/IFNγ+ co-cultures (−20% vs. 100%, p < 0.001). Also, nitrite levels in cultures that were not exposed to neuroinflammation (LPS/IFNγ-) were lower than those in cultures exposed (LPS/IFNγ+; −21% vs. 100%, p < 0.001). The anti-inflammatory cytokine IL10 did not reduce nitrite levels (102% vs. 100%, p > 0.05 compared with LPS/IFNγ+ cultures).

Test compounds

Of the 11 test compounds, 9 (DP1 > BP1 > FBS > FBA > SA1 as well as MLV7, MLV12, MTK, and NVB) reduced nitrite levels at one or more of the concentrations tested.

At the 100-µM concentration, DP1 reduced nitrite concentrations from 100% to 3% (p < 0.001 compared with LPS/IFNγ+). Nitrite levels were comparable with those of positive control, 1400W (3% vs. 0%, p > 0.05). Nitrite levels, however, were higher than in wells with neuronal cells only (BV2−; 3% vs. −20%, p < 0.01) or wells not treated with LPS/IFNγ (LPS/IFNγ-; 3% vs. −21%, p < 0.001). Lower concentrations of DP1 had no favorable effects on nitrite concentrations.

At the 100-µM concentration, BP1 reduced nitrite concentrations from 100% to 28% (p < 0.001 compared with LPS/IFNγ+). This reduction was associated with cell death, however, as demonstrated by both the neuronal viability assay (Fig. 5A) and microscope images captured from cultured cells (Supplementary Fig. S2F).

At the 100-µM concentration, FBS reduced nitrite concentrations from 100% to 38% (p < 0.001 compared with LPS/IFNγ+). Nitrite levels remained higher than those in wells treated with positive control, 1400 W (38% vs. 0%, p < 0.001). Nitrite levels were also higher than those in wells with neuronal cells only (BV2−; 38% vs. −20%, p < 0.001) or wells not treated with LPS/IFNγ (LPS/IFNγ−; 33% vs. −21%, p < 0.001). Lower concentrations of FBS had no favorable effects on nitrite concentrations.

At the 100-µM concentration, FBA reduced nitrite concentrations from 100% to 47% (p < 0.001 compared with LPS/IFNγ+). Nitrite levels remained higher than in wells treated with positive control, 1400W (47% vs. 0%, p < 0.001). Nitrite levels were also higher than those in wells with neuronal cells only (BV2−; 47% vs. −20%, p < 0.001) or wells not treated with LPS/IFNγ (LPS/IFNγ−; 47% vs. −21%, p < 0.001). Lower concentrations of FBA had no favorable effects on nitrite concentrations.

At the 100-µM concentration, SA1 reduced nitrite concentrations from 100% to 64% (p < 0.001 compared with LPS/IFNγ+). Nitrite levels remained higher than those in wells treated with positive control, 1400W (64% vs. 0%, p < 0.001). Nitrite levels were also higher than in wells with neuronal cells only (BV2−; 64% vs. −20%, p < 0.001) or wells not treated with LPS/IFNγ (LPS/IFNγ−; 64% vs. −21%, p < 0.001). Lower concentrations of SA1 had no favorable effects on nitrite concentrations.

The reduction in nitrite concentrations by CA1, MLV7, MLV12, MTK, or NVB was associated with reduced cell viability as observed in the neuronal survival immunoassay and in microscope images captured from cultured cells (Supplementary Figs. S1A–Dand S2K–O). SCH had no effect on nitrate levels (p > 0.05) (Supplementary Fig. S1D).

TNFα levels

The levels of tumor necrosis factor alpha (TNFα) released into the cell culture medium were investigated to assess the anti-inflammatory capacity of the test compounds. The percentage of TNFα in neuronal-BV2 microglial co-cultures exposed to neuroinflammatory mediators LPS and IFNγ (LPS/IFNγ+) was set to 100%. In co-cultures treated with the positive control, anti-inflammatory cytokine IL10, the percentage of TNFα was set to 0%. The anti-inflammatory effect of each test compound was compared with that in co-cultures treated with LPS/IFNγ+ only. Data are summarized in Figure 5C and Supplementary Figure S1E.

Control wells

The TNFα levels in wells with neuronal cells only (BV2−) were lower than those in LPS/IFNγ+ co-cultures (−531% vs. 100%, p < 0.001). Also, TNFα levels in cultures that were not exposed to neuroinflammation (LPS/IFNγ−) were lower than in cultures exposed to LPS/IFNγ+ (−492% vs. 100%, p < 0.001). Neuroprotectant 1400 W also reduced TNFα levels (−342% vs. 100%, p < 0.001) compared with LPS/IFNγ+ cultures).

Test compounds

Of the 11 test compounds, 7 (DP1 > BP1 > FBS > FBA > SA1 as well as MLV7 and MLV12) reduced TNFα levels at one or more of the concentrations tested.

At the 100-µM concentration, DP1 reduced TNFα levels from 100% to −158% (p < 0.001 compared with LPS/IFNγ+). TNFα levels were even lower than in wells treated with positive control IL10 (−158% vs. 0%, p < 0.001). TNFα levels remained higher than those in wells with neuronal cells only (BV2−; −158% vs. −531%, p < 0.001) or wells not treated with LPS/IFNγ (LPS/IFNγ-)(−158% vs. −492%, p < 0.001). Also, a lower concentration of DP1 reduced neuroinflammation (30 µM, −22%, p < 0.01 compared with LPS/IFNγ+).

At the 100-µM concentration, BP1 reduced TNFα levels from 100% to −125%, (p < 0.001 compared with LPS/IFNγ+). The reduced release of TNFα into the culture medium, however, was associated with cell death as demonstrated by both the neuronal viability assay (Fig. 5A) and microscope images captured from cultured cells (Supplementary Fig. S2F). Lower concentrations of BP1 had no favorable effects on nitrite concentrations.

At the 100-µM concentration, FBS reduced TNFα levels from 100% to −158% (p < 0.001 compared with LPS/IFNγ+). TNFα levels were even lower than those in wells treated with positive control, IL10 (−158% vs. 0%, p < 0.001). TNFα levels remained higher than those in wells with neuronal cells only (BV2−; −158% vs. −531%, p < 0.05) or wells not treated with LPS/IFNγ (LPS/IFNγ−; −158% vs. −492%, p < 0.05). A lower concentration of FBS also reduced neuroinflammation (30 µM, 9%, p < 0.05 compared with LPS/IFNγ+).

At the 100-µM concentration, FBA reduced TNFα levels from 100% to −23% (p < 0.01 compared with LPS/IFNγ+). TNFα levels were comparable to those in wells treated with positive control, IL10 (−23% vs. 0%, p > 0.05). TNFα levels remained higher than in wells with neuronal cells only (BV2−; −23% vs. −531%, p < 0.001) or wells not treated with LPS/IFNγ (LPS/IFNγ−; −23% vs. −492%, p < 0.05). Lower concentrations of FBA had no favorable effects on TNFα levels.

At the 100-µM concentration, SA1 reduced TNFα from 100% to −1% (p < 0.05 compared with LPS/IFNγ+). TNFα levels were comparable to those in wells treated with positive control IL10 (−1% vs. 0%, p > 0.05). TNFα levels remained higher than those in wells with neuronal cells only (BV2−; −1% vs. −531% p < 0.05) or wells not treated with LPS/IFNγ (LPS/IFNγ−; −1% vs. −492 −%, p < 0.05). Lower concentrations of SA1 had no favorable effects on TNFα levels.

Co-cultures treated with MLV7 and MLV12 showed a mild favorable effect on TNFα levels (Supplementary Fig. S1E). The reduced TNFα levels, however, were associated with cell death as demonstrated by both the neuronal viability assay and microscope images captured from cultured cells (Supplementary Figs. S1A and S2K, L).

CA1, MTK, NVB, and SCH exposed cultures had low neuronal viability (see representative images in Supplementary Fig. S2), and TNFα levels were therefore not assessed.

Selection of a compound for in vivo validation

The selection of a compound for in vivo validation was primarily based on scoring of the neuroprotective, antioxidant, and anti-inflammatory effects (Table 2, and Supplementary Table S1).^46,47 We also considered the compound toxicity on neuronal and microglial cells by reviewing the microscope images (Supplementary Fig. S2). In addition, we reviewed the available in vitro and in vivo literature on antiseizure effects and toxicity of different compounds (Table 1).

Table 2.

Scoring of the Test Compounds Based on Their In Vitro Effects on Neuronal Viability, Nitrite Production, and TNFα Levels

Compound	Concentration	Neuronal viability	Nitrite	TNFα	Total score
BP1	100 µM	0.15	0.75	1.00	1.90
DP1	100 µM	0.75	0.90	1.00	2.65
FBA	100 nM	0.45	0.45	1.00	1.90
FBS	100 µM	0.75	0.60	1.00	2.35
SA1	100 µM	0.30	0.30	1.00	1.60

Scoring Criteria Are Summarized in Supplementary Table S1.

TNFα, tumor necrosis factor alpha.

Based on in vitro scoring criteria, DP1 (score 2.65 at 100 µM) and FBS (score 2.35 at 100 µM) were the most promising compounds. Visualization of microscope images, however, revealed that FBA-treated cultures (score 1.90) contained a “healthy” cellular profile (Supplementary Fig. S2H) compared with DP1- or FBS-treated cultures (Supplementary Fig. S2G, I). Moreover, available in vivo antiseizure and toxicity data favored FBA over the other compounds.³⁵

Consequently, FBA was selected for in vivo validation to assess whether it would promote structural and functional recovery after LFP-induced TBI in rats.

Effect of FBA on in vivo outcome measures

Impact pressure, apnea time, time to righting, exclusions, and mortality

Table 3 summarizes the impact pressure, apnea time, and time to the righting. The Kruskal–Wallis test revealed a longer apnea time in the TBI-FBApro+ than the TBI-FBApro group (p < 0.05).

Table 3.

Impact Pressure, Postinjury Apnea, and Righting Reflex per Treatment Group

Rx group	Impact pressure (atm)	Apnea (s)	Righting reflex (min)
TBI-VEH (17)	2.9 ± 0.1[2.9, 2.6–3.1]	37.3 ± 16.1[31.0, 22.0–85.0]	17.7 ± 9.7[13.9, 8.0–35.0]
TBI-FBApro (20)	2.9 ± 0.1[2.9, 2.7–3.1]	28.1 ± 10.9[26.5, 8.0–49.0]	12.0 ± 3.3[11.5, 7.2–18.3]
TBI-FBApro+ (10)	2.8 ± 0.2[2.8, 2.5–3.1]	42.9 ± 16.1^* [42.5, 20.0–77.0]	20.1 ± 13.5[16.1, 6.6–48.7]

Data are shown as the mean ± standard deviation of the mean. The number of animals in the given treatment group is in parentheses. Median and range are shown in brackets.

Statistical analysis: *p < 0.05 versus TBI-FBApro.

FBApro, FBA prophylactic treatment; FBApro+, FBA prophylactic treatment plus on-demand doses; TBI, traumatic brain injury; VEH, vehicle treatment.

Acute and follow-up mortality and exclusions are summarized in Figure 2.

Exclusions and acute (<48 h) mortality

In Cohort 1, three rats were excluded (one broken dura, one disconnection of injury cap, and one long [58 min] righting reflex time). There was no acute mortality (0%).

In Cohort 2, one rat was excluded due to a malfunctioning electrode headset. Acute mortality was 9% (3/32).

Follow-up exclusions and mortality

In Cohort 1 and Cohort 2A, there was no follow-up mortality or exclusions for other reasons. In Cohort 2B, follow-up mortality was 15% (2/13); 0% in the TBI-VEHpro+ group and 20% (2/10) in the TBI-FBApro+ group. There were no other follow-up exclusions.

Effect of FBA on the cortical lesion area and plasma pNF-H

Data are summarized in Figure 6.

FIG. 6.

Effect of FBA on the cortical lesion area and plasma pNF-H levels (Cohort 1). (A) In the TBI-FBApro group, the average cortical lesion area (18.9 ± 4.1 mm², median 19.4 mm², range 12.3–25.3 mm²) was smaller than that in the TBI-VEHpro group (24.8 ± 5.7 mm², median 23.5 mm², range 15.4–37.2 mm²)(p < 0.05). Cohen’s delta (1.159) also indicated a strong favorable treatment effect. (B) The average plasma pNF-H levels did not differ between the TBI-FBApro (3288 ± 2396 pg/mL, median 2290 pg/mL, range 1230–8080 pg/mL) and TBI-VEHpro (4970 ± 2716 pg/mL, median 5073 pg/mL, range 2 005–8 363 pg/mL) groups (p > 0.05). Cohen’s delta of 0.657, however, indicated a moderate favorable effect size of FBApro treatment. Open circles in the TBI-VEHpro group indicate the cortical lesion area (panel A) or plasma pNF-H levels (panel B) in vehicle-treated animals from a study that was conducted in parallel with the same tissue and blood sampling timepoints and included in the present analysis (Kajevu et al., 2023⁴⁷; no difference between the studies). Statistical significance: *p < 0.05 compared with the TBI-VEHpro group. Data are presented as box plots with whiskers (minimum and maximum; box: interquartile range; line: median). Each dot shows an experimental value. TBI, traumatic brain injury; VEH, vehicle.

Cortical lesion area

Unfolded cortical maps were prepared to determine the treatment effect on cortical lesion area.⁴⁸ On D14 post-TBI, the average cortical lesion area in the TBI-VEH group was 24.8 ± 5.7 mm² (median 23.5 mm², range 15.4–37.2 mm², and n = 10). In the TBI-FBA group, the cortical lesion area was smaller, 18.9 ± 4.1 mm² (median 19.4 mm², range 12.3–25.3 mm², n = 7, p < 0.05) (Fig. 6A). Cohen’s delta also indicated a strong favorable effect size (1.159).

Plasma levels of pNF-H

The average plasma pNF-H levels at 72 h post-TBI were comparable between the TBI-VEH (4970 ± 2716 pg/mL, median 5073 pg/mL, range 2005–8363 pg/mL, n = 9) and TBI-FBA (3288 ± 2396 pg/mL, median 2290 pg/mL, range 1230–8080 pg/mL, n = 9) groups (p > 0.05) (Fig. 6B). However, Cohen’s delta indicated a moderate favorable effect size of FBA treatment (0.657).

Effect of FBA on acute electrographic seizures during 0–72 h postinjury

As 98% of the acute seizures occur during the 0–72 h postimpact,¹⁴ we focused the analysis on that time period. The occurrence of acute electrographic seizures was comparable between the TBI-VEH groups in Cohorts 1 and 2. Consequently, the data were combined for subsequent analysis.

The occurrence of acute seizures in the TBI-VEH, TBI-FBApro, and TBI-FBApro+ groups is summarized in Figure 7A–D , Figures 8 and 9, and Supplementary Tables S2, S3 and S4.

FIG. 7.

Effect of FBA on acute electrographic seizures. (A) The percentage of rats with electrographic seizures (y-axis) during the 0–72 h post-TBI was comparable between the TBI-VEH (76%), TBI-FBApro (30 mg/kg, 65%), and TBI-FBApro+ (100 mg/kg, 70%) groups (x-axis)(p > 0.05, Fisher’s exact test). The number within each bar shows the number of rats with seizures. (B) The latency to the first electrographic seizure (y-axis) was comparable between the treatment groups (x-axis)(p > 0.05, Kruskal–Wallis). (C) The number of acute electrographic seizures (y-axis) was comparable between the treatment groups (x-axis; p > 0.05, Kruskal–Wallis). (D) The cumulative duration of early seizures (y-axis) varied between the treatment groups (x-axis)(p < 0.05, Kruskal–Wallis). The cumulative seizure duration in the TBI-FBApro (30 mg/kg) and TBI-FBApro+ groups (received an additional dose of 100 mg/kg) was reduced compared with that in the TBI-VEH group (all p < 0.05). Statistical significance: *p < 0.05 compared with the TBI-VEH group. In panels B–D, data are presented as box plots with whiskers (minimum and maximum; box: interquartile range; line: median). Each dot in panels B–D indicates data from one rat. sz, seizure; TBI, traumatic brain injury; TBI-FBApro, rats treated with prophylactic FBA; TBI-FBApro+, rats treated with prophylactic FBA supplemented with additional dose(s) of FBA to suppress status epilepticus; VEH, vehicle (both VEHpro and VEHpro+ animals).

FIG. 8.

Occurrence of acute seizures during the first 0–72 h post-TBI—prophylactic treatment group. (A) TBI-VEHpro group (rats from Cohorts 1 and 2A). Upper panel: A raster plot showing the timing of 201 early seizures, which occurred in 10/14 (71%) rats (y-axis). Lower panel: A histogram showing the number of early seizures per hour during the 72-h epoch. (B) TBI-FBApro group (rats from Cohorts 1 and 2A). Upper panel: A raster plot showing the timing of 136 early seizures, which occurred in 13/20 (65%; rats (y-axis). Lower panel: A histogram showing the number of early seizures per hour during the 72-h epoch. Dashed blue lines divide the 72-h follow-up into 24-h epochs. Purple arrows indicate the times at which prophylactic doses of either VEH or FBA were administered (at 2 and 24 h). sz, seizure; TBI, traumatic brain injury; VEH, vehicle.

FIG. 9.

Occurrence of acute seizures during the first 0–72 h post-TBI — prophylactic treatment group supplemented with on-demand treatment. (A) TBI-VEHpro+ group. Upper panel: A raster plot showing the timing of 57 early seizures, which occurred in 3/3 rats (100%) (y-axis). Lower panel: A histogram showing the number of early seizures per hour during the 72-h epoch. (B) TBI-FBApro+ group. Upper panel: A raster plot showing the timing of 43 seizures, which occurred in 7/10 (70%) rats (y-axis). Lower panel: A histogram showing the number of early seizures per hour during the 72-h epoch. Dashed blue lines divide the 72-h follow-up into 24-h epochs. Purple arrows indicate the time at which the prophylactic doses of either VEH or FBA were administered. Blue triangles indicate the time at which the additional on-demand doses VEH or FBA were administered to suppress ongoing epileptiform activity. Note that the on-demand treatment of FBA (or vehicle) to suppress prolonged epileptiform activity with or without a seizure (ictal-interictal continuum) was mostly administered between 24 and 54 h postinjury; that is, the time period during which 53% of seizures occurred in the vehicle group. TBI, traumatic brain injury; VEH, vehicle.

Seizure occurrence

Fisher’s exact test revealed that the percentage of rats with acute seizures was comparable between the TBI-VEH (76%, 13/17), TBI-FBApro (65%, 13/20), and TBI-FBApro+ (70%, 7/10) groups (p > 0.05) (Fig. 7A).

Latency to the appearance of the first electrographic seizure

In the TBI-VEH group, average latency to the first electrographic seizure was 11 ± 9 h (median 6 h, range 2–30 h). The average latency to the first seizure in the TBI-FBApro (15 ± 14 h, median 7 h, range 2–42 h) and TBI-FBApro+ (19 ± 16 h, median 8 h, range 2–39 h) groups did not differ from that in the TBI-VEH group (both p > 0.05). Also, average latency to the first seizure did not differ between the 2 TBI-FBA treatment groups (p > 0.05) (Fig. 7B).

Number of acute electrographic seizures

Data are summarized in Figure 7C and Supplementary Table S2A, B. In the entire TBI-VEH group, the average seizure number was 15.2 ± 12.4 (median 15.0, range 0.0–40.0). The average seizure number was 6.8 ± 8.6 (median 5.0, range 0.0–33.0) in the TBI-FBApro group and 4.3 ± 4.2 (median 3.5, range 0.0–13.0) in the TBI-FBApro+ group (all comparisons p > 0.05). Cohen’s delta, however, indicated a moderate favorable effect size (0.795) in the TBI-FBApro group and a large favorable effect size (1.058) in the TBI-FBApro+ group compared with vehicle treatment. The number of seizures did not differ between the TBI-FBApro and TBI-FBApro+ groups (p > 0.05). Cohen’s delta showed a small (0.334) favorable treatment effect of the TBI-FBApro+ treatment compared with TBI-FBApro treatment.

When only the rats with seizures were included in the analysis, the Kruskal–Wallis test revealed differences between the treatment groups in the average number of seizures (p < 0.01). The average number of seizures in the TBI-VEH group (n = 13) was 19.9 ± 10.3. The TBI-FBApro (10.5 ± 8.7, n = 13) and TBI-FBApro+ (6.1 ± 3.7, n = 7) groups had a lower average number of seizures than the TBI-VEH group (p < 0.05 and p < 0.01, respectively) (Supplementary Table S2B).

Number of seizures over the 0–72 h follow-up

The evolution of the average number of seizures during the 0–24 h, 25–48 h, and 49–72 h intervals is summarized in Supplementary Table S2A-B. As shown previously, seizure occurrence decayed spontaneously over time (Fig. 8-9) (see TBI-VEH group, Supplementary Table S2A, B).^14,47

On-demand treatment was administered to 1/10 rats between 0–24 h, 7/10 rats between 25–48 h, and 5/10 rats between 49–72 h. Of these animals, 7/10 had seizures during the 0–72 h interval. In the whole animal cohort, analysis of each time interval separately revealed no differences between the TBI-FBApro and TBI-VEH groups (all time intervals p > 0.05). Interestingly, between 49 and 72 h, the TBI-FBApro+ group (i.e., having received additional doses of FBA) had fewer seizures than vehicle-treated animals (0 vs. 2.2 ± 2.8, p < 0.05). Moreover, Cohen’s delta revealed a moderate (0.674) favorable treatment effect between 49 and 72 h by the TBI-FBApro+ treatment over TBI-FBApro treatment (Supplementary Table S2A).

If only the rats with seizures were included in the analysis, average seizure number did not differ between the TBI-FBApro and TBI-VEH groups (all time intervals p > 0.05). Also, the average seizure number did not differ between the TBI-FBApro and TBI-FBApro+ groups (all time intervals p > 0.05). However, Cohen’s delta revealed a moderate favorable effect (0.605) during the 25–48 h interval and a large favorable effect (0.915) during the 49–72 h interval by the TBI-FBApro+ treatment over TBI-FBApro treatment (Supplementary Table S2B).

Average duration of acute electrographic seizures

Data are summarized in Supplementary Table S3A, B. In the whole animal group, the average duration of early seizures occurring during the 0–72 h interval was comparable between the TBI-VEH (46 ± 29 s, median 56 s, range 0–83 s), TBI-FBApro (37 ± 30 s, median 46 s, range 0–82 s), and TBI-FBApro+ (43 ± 41 s, median 42 s, range 0–128 s) groups (p > 0.05). Cohen’s delta revealed a small effect size for the TBI-FBApro group (0.307) and no effect for the TBI-FBApro+ group (0.080) compared with the TBI-VEH group (Supplementary Table S3A). Note that animals in the TBI-FBApro+ group exhibited no seizures during the 49–72 h period.

If only the rats with seizures were included in the analysis, average seizure duration did not differ between the TBI-FBApro, TBI-FBApro+, and TBI-VEH groups (all time intervals p > 0.05). However, Cohen’s delta revealed a small favorable effect (0.418) during the 25–48 h interval and a large favorable effect (1.422) during the 49–72 h interval by the TBI-FBApro+ treatment over the TBI-VEH treatment (Supplementary Table S3B).

Friedman’s 2-way analysis of variance indicated that seizure duration was comparable across all intervals (0–24 h, 25–48 h, and 49–72 h) in the three treatment groups (p > 0.05).

Cumulative duration of acute electrographic seizures

Data are summarized in Figure 7D and Supplementary Table S4A, B. The Kruskal–Wallis test detected differences in the cumulative duration of electrographic seizures among the treatment groups (p < 0.05). In the TBI-VEH group, the average cumulative duration of acute seizures was 896 ± 703 s (median 1145 s, range 0–1833 s). In the TBI-FBApro group, the average cumulative seizure duration was reduced to 368 ± 407 s (median 329 s, range 0–1237 s)(p < 0.05 compared with the TBI-VEH group). In the TBI-FBApro+ group, the cumulative duration of seizures was reduced to 256 ± 327 s (median 218 s, range 0–1108 s; p < 0.05). Cohen’s delta suggested a large favorable treatment effect size in both the TBI-FBApro (0.941) and TBI-FBApro+ (1.076) treatment groups compared with the TBI-VEH treatment group.

When only the rats with seizures were included in the analysis, the Kruskal–Wallis test detected differences in the cumulative duration of electrographic seizures among the treatment groups (p < 0.01) (Supplementary Table S4B). In the TBI-VEH group (n = 13), the average cumulative seizure duration was 1 172 ± 556 s. In the TBI-FBApro (566 ± 376 s, n = 13) and TBI-FBApro+ (365 ± 337 s, n = 7; p < 0.01) groups, the average cumulative seizure duration was reduced compared with that in the TBI-VEH group (both p < 0.01). Cohen’s delta suggested a large favorable treatment effect size in both the TBI-FBApro (1.279) and TBI-FBApro+ (1.635) treatment groups compared with the TBI-VEH treatment group. Also, Cohen’s delta indicated a moderate effect of FBApro+ treatment compared with FBApro treatment (0.551). Cohen’s delta indicated moderate to strong effect of TBI-FBA treatment over TBI-VEH treatment at all time intervals. Moreover, during the 25–48 h interval, Cohen’s delta revealed a large favorable effect (0.836) by TBI-FBApro treatment over TBI-VEH treatment.

Effect of on-demand FBA administration on relative EEG power

Next, we assessed the electrographic signature of FBA treatment on the relative power of the delta, theta, alpha, beta, and gamma bands in EEG. Data are summarized in Supplementary Figure S3. The EEG epochs were sampled 60 min before and 60 min after the first dose of FBA on-demand treatment and compared, matching the expected increase of FBA brain concentration after the on-demand administration.

In the TBI-VEHpro+ group, average relative power of all EEG bands before and after vehicle administration was comparable in all brain regions (C3, C4, O1, and O2; p > 0.05 Wilcoxon’s test).

In the TBI-FBApro+ group, the first dose of on-demand treatment (100 mg/kg, i.p.) reduced average relative theta power in the ipsilateral frontal area (C3; 22% vs. 20%, p < 0.05). In parallel, average alpha power decreased ipsilaterally in both C3 (7% vs. 5%, p < 0.01) and O1 (6% vs. 5%, p < 0.05) as well as contralaterally in both C4 (10% vs. 7%, p < 0.01) and O2 (9% vs. 7%, p < 0.01). The average gamma power increased in the occipital area in both O1 (6% vs. 11%, p < 0.01) and O2 (12% vs. 16%, p < 0.05). No effects on average delta or beta power were detected.

Discussion

The present study investigated whether antiseizure compounds discovered through structure-based designing could be repurposed to mitigate brain damage and alleviate early seizures after TBI. We hypothesized that a compound capable of promoting neuronal viability, reducing neuroinflammation and nitrite-induced oxidative stress in vitro, would have a favorable disease-modifying effect in a rat model of TBI. Based on a previously defined pipeline,⁴⁶ we tested the effects of the compounds in vitro prior to validating the most promising compounds in vivo. We had four major findings: (i) Of the 11 computer-aided identified compounds, 3 compounds, DP1, FBA, and FBS, exhibited a promising in vitro therapeutic effect. The remaining compounds either exhibited some toxicity (BP1, MTK, and SA1) or had no effect in promoting neuronal survival CA1, MLV7, MLV12, NVB, or SCH. (ii) FBA, a compound with a promising in vitro therapeutic effect, reduced the cortical lesion area at 2 weeks post-TBI. (iii) FBA also had a promising antiseizure effect during the acute post-TBI phase. (iv) Administration of an additional dose of FBA on-demand if epileptiform activity emerged showed additive antiseizure effects.

DP1, FBS, and FBA had the most promising in vitro therapeutic profile

Existing evidence indicates that TBI results in neuronal cell death and a cascade of molecular changes, including neuroinflammation and oxidative stress.^5,6,49 Several studies report that treatments enhancing neuronal survival and alleviating neuroinflammation and oxidative stress could promote recovery following TBI.^5,19 Consequently, the ability of structure-based design compounds to enhance neuronal viability and reduce neuroinflammation and oxidative stress was tested in vitro before proceeding to in vivo validation. A score of in vitro outcome measures^46,47 was used to determine the compound with the most promising treatment effects. In addition, visualization of microscope images enabled us to perform a qualitative analysis of the cell cultures to avoid false-positive results based on in vitro tests only. Further, preliminary in vivo data on antiseizure efficacy in mouse electroshock model³⁵ were considered before selecting the compound for in vivo testing.

Compound DP1 (donepezil) is an acetylcholinesterase inhibitor used to treat mild and moderate Alzheimer’s disease.⁵⁰ Previous data revealed that 0.1–10 µM DP1 confers neuroprotection against oxygen–glucose deprivation and NMDA excitotoxicity in cortical neurons, as well as protection against amyloid-β toxicity in septal neurons.³² In another study, 1–50 µM of DP1 reduced proinflammatory mRNA levels in BV2 microglial cells. In the same study, 1 mg/kg of DP1 reduced: (a) amyloid-β-induced inflammation in a mouse model of Alzheimer’s disease and (b) LPS-induced COX2 and IL6 levels in wild-type mice.⁵¹ In addition, DP1 had an antiseizure effect in a mouse model of Dravet syndrome.⁵² Together, these studies provided evidence of the neuroprotective and antiseizure effects of DP1, and thus we tested the effects of DP1 in our neuronal-BV2 microglial co-culture model. In our study, quantitative analysis using a MAP2-immunoassay showed that 1–100 µM of DP1 had a neuroprotective effect in co-cultures. Some cellular degeneration was observed in co-cultures treated with 100 µM of DP1, however, suggesting that higher doses of DP1 could be cytotoxic. Overall, our in vitro data are consistent with previous data showing that DP1 has neuroprotective properties at lower concentrations.

The FBS compound (3-p-fluorbenzyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-one-2,2-dioxide) was designed to block sodium 1.2 voltage-gated channels (Nav1.2).³⁵ FBS demonstrated potent Nav1.2 blocking capacity in docking simulations. Also, 100 µM of FBS inhibited Nav1.2 currents in HEK 293 cell patch clamp experiments. FBS had only a weak antiseizure effect, however, in a mouse maximal electroshock model.³⁵ In previous reports, sodium channel blockers demonstrated neuroprotective effects both in vitro and in vivo.^53,54 Sodium channel blockers prevent the propagation of action potentials, thus mitigating excitotoxicity, which could account for their neuroprotective capacity.⁵⁵ Therefore, we tested the neuroprotective effects of FBS in our co-culture model. In our in vitro investigations, a MAP2-immunoassay showed that treatment with FBS at 30 µM and 100 µM has promising neuroprotective effects. Some cellular degeneration was observed in co-cultures treated with 100 µM FBS, however, suggesting that higher doses of FBS could be cytotoxic.

Further analysis involving the allocation of a score based on the effect of a compound on in vitro outcome measures revealed that DP1 and FBS had the highest overall scores. Because some cellular degeneration was observed in co-cultures treated with the highest concentration (100 µM) of DP1 or FBS, these compounds were excluded from further in vivo analysis. Whether this mild discrepancy between the data obtained in the immunoassay and microscopy related to different assay sensitivities remains to be studied. Nevertheless, this observation highlights the importance of microscope imaging as an additional investigation to assess the in vitro effect or toxicity of experimental compounds. We attributed the low nitrite and TNFα levels observed in co-cultures treated with 100 µM of DP1 or FBS to the reduced survival of both neuronal and microglial cells observed in the microscope images. Overall, as both DP1 and FBS demonstrated some degree of cytotoxicity, they were not selected for further in vivo testing.

The compound FBA (N-[p-fluoro] benzyl-2-hydroxy isobutylamide) was designed and synthesized as an antiseizure drug.³⁵ In both docking simulations and in HEK 293 cell culture investigations, FBA did not demonstrate an ability to block Nav1.2 channels. In vivo assays in a mouse maximal electroshock model, however, showed that FBA has potent antiseizure activity.³⁵ The mechanism underlying the antiseizure activity of FBA is still under investigation. The present analysis of data from the MAP2-immunoassay and microscope images revealed that treatment with 100 µM of FBA had the most promising neuroprotective effect. Moreover, 100 µM of FBA had promising antioxidant (low nitrite levels) and anti-inflammatory (low TNFα levels) effects. The promising treatment effect of FBA in our in vitro study and in the mouse maximal electroshock model³⁵ prompted us to select FBA for further in vivo testing in the LFP-induced TBI model used in our laboratory. The in vitro validation experiment was conducted by an individual blinded to the antiseizure effects exhibited by FBA.

FBA mitigated brain damage by reducing the cortical lesion area

Within seconds to minutes following TBI, a myriad of secondary molecular modifications occur, resulting in cellular and axonal damage.^5,7,56 Our previous magnetic resonance imaging follow-up study showed that the cortical lesion area expands rapidly during the first 3 weeks post-TBI.⁵⁶ Data from many other laboratories indicate that early treatment with neuroprotective compounds reduces the post-TBI lesion size.^57
–59 In our study, cortical lesion area was assessed on D14 post-TBI. FBA treatment administered at 2 and 24 h postimpact reduced the cortical lesion area by 24%. Although the neuroprotective mechanism targeted by FBA and the duration of the postinjury expression remain to be investigated, our data suggest a relatively wide therapeutic time window (hours rather than minutes) for treatment initiation. Moreover, we found a moderate favorable FBA treatment effect on the axonal injury marker pNF-H, supporting the in vitro-predicted neuroprotective effect of FBA in vivo.

FBA had a promising effect on acute post-TBI seizures

Previous reports indicate the importance and efficacy of antiseizure prophylaxis for the treatment of acute seizures post-TBI in humans and rodents.^60
–62 In electrical stimulation or chemoconvulsant models of SE, prevention or reduction of early seizures mitigated secondary brain damage and epileptogenesis.⁶³Although the long-term benefits of antiseizure prophylaxis after TBI are still unclear,^64,65 evidence suggests that patients with immediate or early seizures have a higher risk of developing PTE, poor cognitive recovery, and an elevated risk of death.^8,12

To prevent acute seizures post-TBI, FBA was administered at 2 and 24 h after impact as a prophylactic treatment. Moreover, to mimic the clinical treatment of SE, one subcohort of animals received additional doses of FBA on-demand if epileptiform activity lasting at least 5 min emerged during the online follow-up despite prophylactic treatment.^66,67 Although prophylactic FBA treatment alone or FBA treatment complemented with on-demand administration exhibited no statistically significant reduction in the average seizure number, Cohen’s delta suggested a large favorable treatment effect in this relatively small cohort size. The effect was even clearer when the analysis was performed on rats with seizures only. The reduction in the seizure number by prophylactic FBA was augmented by administering additional on-demand FBA treatment. As expected, the reduction in the seizure number was associated with a remarkable reduction (∼60%) in cumulative seizure duration.

Previous investigations indicate that FBA has a peak effect in mice at 30 min after administration.³⁵ Based on this preliminary data, the acute doses of FBA were separated by at least 30 min to allow the compound to reach its maximum brain concentration before administering a second dose. As FBA has a short half-life, it might be necessary to administer more than two acute doses of FBA to achieve its full antiseizure potential. These additional doses could confer some toxicity, however, as suggested by a 20% follow-up mortality in the subgroup of animals that received additional doses of FBA. Future studies are warranted to investigate the full spectrum of adverse effects of FBA.

Relative power analysis revealed an increase in relative gamma power bilaterally in the occipital areas. Previous studies revealed that increased gamma power indicates the activation of parvalbumin-positive inhibitory GABAergic neurons.^68
–70 Functionally, increased gamma power is associated with improved attention, cognition, and working memory.^70
–72 It is also possible that the increased gamma power is related to increased awareness and attention triggered by the intraperitoneal injection, even though the rats were undergoing electrographic SE. The increase in the relative gamma power was more prominent after FBA injection than after vehicle injection, however, favoring a drug effect rather than an injection effect. We also found that FBA reduced relative theta power in the ipsilateral frontal area. The significance of the relative theta power is still under investigation, but some reports suggest that reduced theta power correlates with improved cognition.^73,74 FBA also induced a reduction in relative alpha power at all recording sites. Previous studies on the functional significance of low alpha power are inconclusive.⁷⁵ One clinical study suggests that low alpha power could indicate the presence of axonal injury and loss of white matter.⁷⁶ In the present study, however, FBA treatment was associated with reduced, rather than increased, levels of the axonal injury marker pNF-H. Other studies in humans report that reduced alpha rhythm is associated with cognitive impairment and poor quality of the N1 and N2 stages of sleep.^77,78 The functional significance of a drug-induced reduction in relative alpha power in rats remains to be further explored.

Taken together, these data suggest that FBA is a promising compound for the treatment of early seizures, both as a prophylactic treatment started soon after TBI diagnosis and as an on-demand treatment to be administered when epileptiform activity emerges. Further studies are needed to maximize the antiseizure efficacy and tolerability of FBA.

Caveats

Only FBA was tested in vivo, even though also DP1 and FBS were promising based on in vitro testing of neuroprotection and measurement of levels of anti-inflammatory and antioxidant markers. Our final decision to proceed to proof-of-concept in vivo trial took into account also the qualitative analysis of cell death detected on microscope images. As the microscopic analysis was performed at 24 h posttreatment timepoint only, further safety studies are warranted to explore the safety profile of all top-scoring compounds in detail to consider their evaluation in vivo.

FBA on-demand treatment (FBApro+) was only given to non-responding rats in the FBApro group, that is, to rats that showed seizures despite prophylactic FBA treatment. As controls, we investigated vehicle-treated rats that had been pretreated with vehicle to control the possible injection/stress effect on epileptiform activity. It remains to be explored in a larger animal cohort (including both males and females), whether the administration of FBA will show antiseizure efficacy as an on-demand administration to subjects undergoing SE with or without prior clinically used antiseizure treatments.

Conclusion

Our data demonstrate that computer-guided identification of antiseizure compounds is a viable approach for compound discovery to treat post-TBI epileptiform activity. FBA, a computer-aided designed compound with preliminary evidence of antiseizure activity in a mouse model exhibited antiseizure and neuroprotective efficacy after TBI. Further studies in larger animal cohorts with both sexes and multiple dosing are needed to reveal the full potential of FBA as an antiseizure treatment.

Transparency, Rigor, and Reproducibility Statement

The study designs and analytic plan were not formally preregistered on any repository. Computer-assisted drug discovery approaches were used to identify and select 11 antiseizure compounds based on their molecular structure. The neuroprotective potential of the test compounds was investigated in vitro in neuronal-BV2 microglial co-cultures. After a 48-h exposure to test compounds, neuronal viability (MAP2 immunoassay), nitrite, and TNFα assays were performed. FBA, a compound which exhibited strong antiseizure efficacy in a mouse maximal electroshock model, had the most promising therapeutic profile in vitro and was selected for validation in a rat model of LFP-induced TBI. In total, 54 rats were injured using LFP, 4 were excluded (1 for prolonged recovery post-TBI, 1 for disconnected injury cap, 1 for broken dura, and 1 for problem with headset connection). Acute mortality was 6% (3/54). Using excel, the remaining 47 rats were randomly assigned to the following treatment groups: TBI-vehicle (TBI-VEH, n = 17 [TBI-VEHpro, n = 14 and TBI-VEHpro+, n = 3]), TBI-FBA prophylactic (TBI-FBApro, n = 20), and TBI-FBA prophylactic, plus additional FBA on-demand treatment (TBI-FBApro+, n = 13). Follow-up mortality was 4% (2/54). Prophylactic treatment was administered at 2 and 24 h after injury. In addition to the prophylactic treatment rats in the TBI-VEHpro+ or TBI-FBApro+ group received up to two doses of on-demand treatment if we observed any SE-related epileptiform activity. The experiment was replicated in two separate cohorts. The effect of FBA on in vivo outcome was assessed through the analysis of seizure occurrence and EEG power spectrum, extent of cortical lesion area, and plasma levels of pNF-H. The in vitro study was performed by an individual blinded to the antiseizure of FBA. In vivo data analysis was conducted in a blinded manner. As the occurrence of seizures between the TBI-VEH group in Cohort 1 was comparable to that in Cohort 2, we concluded that there was consistency between the cohorts. The normality of the data was assessed using the Shapiro-Wilkinson test. A p value <0.05 was considered as significant. All equipment and analytical reagents used to perform experiments or data analysis are available from the sources indicated in the article.

Footnotes

Acknowledgments

The authors thank Jarmo Hartikainen, Merja Lukkari, and Petra Mäkinen for their excellent technical assistance.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are not publicly available but could be acquired from the corresponding author on reasonable request.

Authors’ Contributions

Conception and design: N.K., I.B., P.A., A.L., A.T., and A.P. Methodology and investigation: N.K., I.B., P.A., A.L., L.S., M.L.V., M.C., T.N., and E.H. Data analysis: N.K., I.B., P.A., A.L., and A.P. Article writing with input from all authors: N.K., I.B., P.A., A.L., N.P., L.G., A.T., and A.P. Funding acquisition: N.K., M.H., L.G., A.T., and A.P.

Author Disclosure Statement

The authors have no competing interest to disclose.

Funding Information

This study was supported by the GenomMed Horizon 2020 Framework Programme of the European Union (Marie Sklodowska-Curie grant agreement no. 740264), the Medical Research Council of the Academy of Finland (Grant nos. 272249, 273909, 317203, and 338182), the Sigrid Juselius Foundation and the Strategic Neuroscience Funding of the University of Eastern Finland, and the ApoyoDravet and the Agencia Nacional de Promoción de la Investigación, el Desarrollo Tecnológico y la Innovación (Agencia I + D+i, grants PICT 2019-1075, PICT-2020-02077) and the Universidad Nacional de La Plata, Argentina (Project no. x-883).

Supplementary Material

Supplementary Data

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

References

Dewan

, Rattani

, Gupta

, et al. Estimating the global incidence of traumatic brain injury. J Neurosurg, 2019; 130(4):1080–1097; doi: 10.3171/2017.10.JNS17352

Maas

AIR

, Menon

, Manley

, et al.; InTBIR Participants and Investigators. Traumatic brain injury: Progress and challenges in prevention, clinical care, and research. Lancet Neurol, 2022; 21(11):1004–1060; doi: 10.1016/S1474-4422(22)00309-X

Werner

, Engelhard

. Pathophysiology of traumatic brain injury. Br J Anaesth, 2007; 99(1):4–9; doi: 10.1093/bja/aem131

Vella

, Crandall

, Patel

. Acute management of traumatic brain injury. Surg Clin North Am, 2017; 97(5):1015–1030; doi: 10.1016/j.suc.2017.06.003

Jarrahi

, Braun

, Ahluwalia

, et al. Revisiting traumatic brain injury: From molecular mechanisms to therapeutic interventions. Biomedicines, 2020; 8(10):1–42; doi: 10.3390/biomedicines8100389

Pitkänen

, Kharatishvili

, Karhunen

, et al. Epileptogenesis in experimental models. Epilepsia, 2007; 48(Suppl. 2):13–20; doi: 10.1111/j.1528-1167.2007.01063.x

Pitkänen

, Kemppainen

, Ndode-Ekane

, et al. Posttraumatic epilepsy—disease or comorbidity? Epilepsy Behav, 2014; 38:19–24; doi: 10.1016/j.yebeh.2014.01.013

Laing

, Gabbe

, Chen

, et al. Risk factors and prognosis of early posttraumatic seizures in moderate to severe traumatic brain injury. JAMA Neurol, 2022; 79(4):334–341; doi: 10.1001/jamaneurol.2021.5420

Vespa

, Nuwer

, Nenov

, et al. Increased incidence and impact of nonconvulsive and convulsive seizures after traumatic brain injury as detected by continuous electroencephalographic monitoring. J Neurosurg, 1999; 91(5):750–760; doi: 10.3171/jns.1999.91.5.0750

10.

Vespa

, Miller

, McArthur

, et al. Nonconvulsive electrographic seizures after traumatic brain injury result in a delayed, prolonged increase in intracranial pressure and metabolic crisis. Crit Care Med, 2007; 35(12):2830–2836; doi: 10.1097/01.CCM.0000295667.66853.BC

11.

Vespa

, McArthur

, Xu

, et al. Nonconvulsive seizures after traumatic brain injury are associated with hippocampal atrophy. Neurology, 2010; 75(9):792–798; doi: 10.1212/WNL.0b013e3181f07334

12.

Ritter

, Wagner

, Fabio

, et al. Incidence and risk factors of posttraumatic seizures following traumatic brain injury: A traumatic brain injury model systems study. Epilepsia, 2016; 57(12):1968–1977; doi: 10.1111/epi.13582

13.

Annegers

, Grabow

, Groover

, et al. Seizures after head trauma. Neurology, 1980; 30(7 Pt 1):683–689; doi: 10.1212/WNL.30.7.683

14.

Andrade

, Banuelos-Cabrera

, Lapinlampi

, et al. Acute Non-Convulsive status epilepticus after experimental traumatic brain injury in rats. J Neurotrauma, 2019; 36(11):1890–1907; doi: 10.1089/neu.2018.6107

15.

Annegers

, Hauser

, Coan

, et al. A population-based study of seizures after traumatic brain injuries. N Engl J Med, 1998; 338(1):20–24; doi: 10.1056/NEJM199801013380104

16.

Englander

, Bushnik

, Duong

, et al. Analyzing risk factors for late posttraumatic seizures: A prospective, multicenter investigation. Arch Phys Med Rehabil, 2003; 84(3):365–373; doi: 10.1053/apmr.2003.50022

17.

Schouten

. Neuroprotection in traumatic brain injury: A complex struggle against the biology of nature. Curr Opin Crit Care, 2007; 13(2):134–142; doi: 10.1097/MCC.0b013e3280895d5c

18.

Dulla

, Pitkänen

. Novel approaches to prevent epileptogenesis after traumatic brain injury. Neurotherapeutics, 2021; 18(3):1582–1601; doi: 10.1007/s13311-021-01119-1

19.

Golub

, Reddy

. Post-Traumatic epilepsy and comorbidities: Advanced models, molecular mechanisms, biomarkers, and novel therapeutic interventions. Pharmacol Rev, 2022; 74(2):387–438; doi: 10.1124/pharmrev.121.000375

20.

Talevi

. Computer-Aided drug discovery and design: Recent advances and future prospects. Methods Molecular Biol Humana Press Inc, 2024:1–20; doi: 10.1007/978-1-0716-3441-7_1

21.

Djikic

, Martí

, Spyrakis

, et al. Human dopamine transporter: The first implementation of a combined in silico/in vitro approach revealing the substrate and inhibitor specificities. J Biomol Struct Dyn, 2019; 37(2):291–306; doi: 10.1080/07391102.2018.1426044

22.

Chandra

, Wang

, Tao

, et al. Computer-aided discovery of a new nav1.7 inhibitor for treatment of pain and itch. Anesthesiology, 2020; 133(3):611–627; doi: 10.1097/ALN.0000000000003427

23.

Kaldor

, Kalish

, Davies

, et al. Viracept (Nelfinavir Mesylate, AG1343): A potent, orally bioavailable inhibitor of HIV-1 protease. J Med Chem, 1997; 40(24):3979–3985; doi: 10.1021/jm9704098

24.

Pun

, Ozerov

, Zhavoronkov

. AI-Powered therapeutic target discovery. Trends Pharmacol Sci, 2023; 44(9):561–572; doi: 10.1016/j.tips.2023.06.010

25.

Lipponen

, Natunen

, Hujo

, et al. In vitro and in vivo pipeline for validation of disease-modifying effects of systems biology-derived network treatments for traumatic brain injury—lessons learned. Int J Mol Sci, 2019; 20(21); doi: 10.3390/ijms20215395

26.

Kajevu

, Lipponen

, Andrade

, et al. Treatment of status epilepticus after traumatic brain injury using an antiseizure drug combined with a tissue recovery enhancer revealed by systems biology. Int J Mol Sci, 2023; 24(18); doi: 10.3390/ijms241814049

27.

Sabatier

, Palestro

, Enrique

, et al. Design, synthesis and biological evaluation of N-substituted α-hydroxyimides and 1,2,3-oxathiazolidine-4-one-2,2-dioxides with anticonvulsant activity. J Enzyme Inhib Med Chem, 2019; 34(1):1465–1473; doi: 10.1080/14756366.2019.1651722

28.

Ekolle Ndode-Ekane

, Kharatishvili

, Pitkänen

. Unfolded maps for quantitative analysis of cortical lesion location and extent after traumatic brain injury. J Neurotrauma, 2017; 34(2):459–474; doi: 10.1089/neu.2016.4404

29.

Boone

, Weisz

, Willey

, et al. Traumatic brain injury induces long-lasting changes in immune and regenerative signaling. PLoS One, 2019; 14(4):e0214741; doi: 10.1371/journal.pone.0214741

30.

Sugimoto

, Ogura

, Arai

, et al. REVIEW-New drug and recent technique-research and development of donepezil hydrochloride, a new type of acetylcholinesterase inhibitor. Jpn J Pharmacol, 2002.

31.

Akasofu

, Kimura

, Kosasa

, et al. Study of neuroprotection of donepezil, a therapy for Alzheimer’s disease. Chem Biol Interact, 2008; 175(1–3):222–226; doi: 10.1016/j.cbi.2008.04.045

32.

Kim

, Lee

, Park

, et al. Donepezil regulates LPS and aβ-stimulated neuroinflammation through MAPK/NLRP3 inflammasome/STAT3 signaling. Int J Mol Sci, 2021; 22(19); doi: 10.3390/ijms221910637

33.

Wong

, Thelin

, Escayg

. Donepezil increases resistance to induced seizures in a mouse model of Dravet syndrome. Ann Clin Transl Neurol, 2019; 6(8):1566–1571; doi: 10.1002/acn3.50848

34.

Shimojo

, Takasugi

, Yamamoto

, et al. Neuroprotective action of a novel compound-M50463-in primary cultured neurons 1. Brain Res, 1999; 815(1):131–139

35.

Kamisli

, Basaran

, Batcioglu

, et al. Neuroprotective effects of the new Na channel blocker rs100642 in global ischemic brain injury. Arch Med Sci, 2019; 15(2):467–474; doi: 10.5114/aoms.2017.72550

36.

Madge

. Chapter 6. Sodium Channels: Recent Developments and Therapeutic Potential. Annual Reports in Medicinal Chemistry, 1998; 33:51–60; doi: 10.1016/S0065-7743(08)61071-4In: Academic Press Inc.;

37.

Immonen

, Kharatishvili

, Niskanen

, et al. Distinct MRI pattern in lesional and perilesional area after traumatic brain injury in rat - 11 months follow-up. Exp Neurol, 2009; 215(1):29–40; doi: 10.1016/j.expneurol.2008.09.009

38.

Homsi

, Piaggio

, Croci

, et al. Blockade of acute microglial activation by minocycline promotes neuroprotection and reduces locomotor hyperactivity after closed head injury in mice: A Twelve-Week follow-up study. J Neurotrauma, 2010; 27(5):911–921; doi: 10.1089/neu.2009.1223

39.

Yang

, Liu

, Ren

, et al. DRα1-MOG-35-55 treatment reduces lesion volumes and improves neurological deficits after traumatic brain injury. Metab Brain Dis, 2017; 32(5):1395–1402; doi: 10.1007/s11011-017-9991-6

40.

Williams

, Bhatti

, Brown

, et al. Early Single-Dose treatment with exosomes provides neuroprotection and improves blood-brain barrier integrity in swine model of traumatic brain injury and hemorrhagic shock. J Trauma Acute Care Surg, 2020; 88(2):207–218; doi: 10.1097/TA.0000000000002563

41.

Gultekin

, Huang

, Clavisi

, et al. Pharmacological interventions in traumatic brain injury: Can we rely on systematic reviews for evidence? Injury, 2016; 47(3):516–524; doi: 10.1016/j.injury.2015.10.011

42.

Mountney

, Shear

, Potter

, et al. Ethosuximide and phenytoin dose-dependently attenuate acute nonconvulsive seizures after traumatic brain injury in rats. J Neurotrauma, 2013; 30(23):1973–1982; doi: 10.1089/neu.2013.3001

43.

Schierhout

, Roberts

. Prophylactic antiepileptic agents after head injury: A systematic review. J Neurol Neurosurg Psychiatry, 1998; 64(1):108–112; doi: 10.1136/jnnp.64.1.108

44.

Pitkänen

. Therapeutic approaches to epileptogenesis - Hope on the Horizon. In: Epilepsia, 2010; 51(Suppl 3):2–17; doi: 10.1111/j.1528-1167.2010.02602.x

45.

Hernandez

. Preventing post-traumatic epilepsy after brain injury: Weighing the costs and benefits of anticonvulsant prophylaxis. Trends Pharmacol Sci, 1997; 18(2):59–62; doi: 10.1016/S0165-6147(97)89801-X

46.

Coelho

LMG

, Blacker

, Hsu

, et al. Association of early seizure prophylaxis with posttraumatic seizures and mortality. Neur Clin Pract, 2023; 13(3); doi: 10.1212/cpj.0000000000200145

47.

Trinka

, Leitinger

. Management of status epilepticus, refractory status epilepticus, and super-refractory status epilepticus. Continuum (Minneap Minn), 2022; 28(2):559–602

48.

Migdady

, Rosenthal

, Cock

. Management of status epilepticus: A narrative review. Anaesthesia, 2022; 77(S1):78–91; doi: 10.1111/anae.15606

49.

Sohal

, Zhang

, Yizhar

, et al. Parvalbumin neurons and gamma rhythms enhance cortical circuit performance. Nature, 2009; 459(7247):698–702; doi: 10.1038/nature07991

50.

Guerriero

, Giza

, Rotenberg

. Glutamate and GABA imbalance following traumatic brain injury. Curr Neurol Neurosci Rep, 2015; 15(5):27; doi: 10.1007/s11910-015-0545-1

51.

Ferguson

, Glick

, Huguenard

. Prefrontal PV interneurons facilitate attention and are linked to attentional dysfunction in a mouse model of absence epilepsy. Elife, 2023; 12; doi: 10.7554/eLife.78349

52.

Álvarez

, Sampedro

, Figueroa

, et al. Reductions in qEEG slowing over 1 year and after treatment with Cerebrolysin in patients with moderate-severe traumatic brain injury. J Neural Transm (Vienna), 2008; 115(5):683–692; doi: 10.1007/s00702-008-0024-9

53.

Cho

KKA

, Hoch

, Lee

, et al. Gamma rhythms link prefrontal interneuron dysfunction with cognitive inflexibility in dlx5/6+/- mice. Neuron, 2015; 85(6):1332–1343; doi: 10.1016/j.neuron.2015.02.019

54.

Prichep

. Quantitative EEG and Electromagnetic Brain Imaging in aging and in the Evolution of Dementia. Annals of the New York Academy of Sciences Blackwell Publishing Inc.; 2007; pp. 156–167; doi: 10.1196/annals.1379.008

55.

Cao

, Han

, Peng

, et al. Correlation between resting theta power and cognitive performance in patients with schizophrenia. Front Hum Neurosci, 2022; 16:853994; doi: 10.3389/fnhum.2022.853994

56.

Nuwer

, Hovda

, Schrader

, et al. Routine and quantitative EEG in mild traumatic brain injury. Clin Neurophysiol, 2005; 116(9):2001–2025; doi: 10.1016/j.clinph.2005.05.008

57.

Valdés-Hernández

, Ojeda-González

, Martínez-Montes

, et al. White matter architecture rather than cortical surface area correlates with the EEG alpha rhythm. Neuroimage, 2010; 49(3):2328–2339; doi: 10.1016/j.neuroimage.2009.10.030

58.

Klimesch

. EEG alpha and theta oscillations reflect cognitive and memory performance: A review and analysis. Brain Res Brain Res Rev, 1999; 29(2–3):169–195; doi: 10.1016/S0165-0173(98)00056-3

59.

Zhao

, Chen

, Duan

, et al. Impaired alpha rhythm in bad sleeper during nap: A high-density electroencephalogram study. J Sleep Res, 2022; 31(1):e13419; doi: 10.1111/jsr.13419

60.

Maier

, Lampel

, Bhutani

, et al. Hand dermatitis: A focus on allergic contact dermatitis to biocides. Dermatol Clin, 2009; 27(3):251–264; doi: 10.1016/j.det.2009.05.007

61.

Juliano

, Magrini

. Cosmetic ingredients as emerging pollutants of environmental and health concern. A mini-review. Cosmetics, 2017; 4(2):11; doi: 10.3390/cosmetics4020011

62.

Enrique

, Martín

, Sbaraglini

, et al. Parabens inhibit hNaV 1.2 channels. Biomed Pharmacother, 2020; 128:110250; doi: 10.1016/j.biopha.2020.110250

63.

Talevi

, Enrique

, Bruno-Blanch

. Anticonvulsant activity of artificial sweeteners: A structural link between sweet-taste receptor T1R3 and brain glutamate receptors. Bioorg Med Chem Lett, 2012; 22(12):4072–4074; doi: 10.1016/j.bmcl.2012.04.076

64.

Van Eyk

. The effect of five artificial sweeteners on Caco-2, HT-29 and HEK-293 cells. Drug Chem Toxicol, 2015; 38(3):318–327; doi: 10.3109/01480545.2014.966381

65.

Pisera-Fuster

, Otero

, Talevi

, et al. Anticonvulsant effect of sodium cyclamate and propylparaben on pentylenetetrazol-induced seizures in zebrafish. Synapse, 2017; 71(4); doi: 10.1002/syn.21961

66.

Kimura

, Hayashida

, Eisenach

, et al. Relief of hypersensitivity after nerve injury from systemic donepezil involves spinal cholinergic and γ-Aminobutyric acid mechanisms. Anesthesiology, 2013; 118(1):173–180; doi: 10.1097/ALN.0b013e318277a81c

67.

Couyoupetrou

, Talevi

, Me

, et al. Latin American congress on epilepsy—Abstracts. Epileptic Disorders, 2014; 16(NS3):1–54; doi: 10.1684/epd.2014.0694

68.

Villalba

, Enrique

, Higgs

, et al. Novel sulfamides and sulfamates derived from amino esters: Synthetic studies and anticonvulsant activity. Eur J Pharmacol, 2016; 774:55–63; doi: 10.1016/j.ejphar.2016.02.001

69.

Villalba

, Palestro

, Ceruso

, et al. Sulfamide derivatives with selective carbonic anhydrase VII inhibitory action. Bioorg Med Chem, 2016; 24(4):894–901; doi: 10.1016/j.bmc.2016.01.012

70.

Bousquet

, Demoly

, Humbert

. Montelukast in guidelines and beyond. Adv Ther, 2009; 26(6):575–587; doi: 10.1007/s12325-009-0038-1

71.

Al Saadi

, Meo

, Mustafa

, et al. Effects of montelukast on free radical production in whole blood and isolated human Polymorphonuclear Neutrophils (PMNs) in asthmatic children. Saudi Pharm J, 2011; 19(4):215–220; doi: 10.1016/j.jsps.2011.06.002

72.

Cevik

, Solmaz

, Aksoy

, et al. Montelukast inhibits pentylenetetrazol-induced seizures in rats. Med Sci Monit, 2015; 21:869–874; doi: 10.1265/9/MSM.892932

73.

Llanos

, Enrique

, Sbaraglini

, et al. Structure-Based virtual screening identifies novobiocin, montelukast, and cinnarizine as TRPV1 modulators with anticonvulsant activity in vivo . J Chem Inf Model, 2022; 62(12):3008–3022; doi: 10.1021/acs.jcim.2c00312

74.

Rodríguez-Cerrato

, del Prado

, Huelves

, et al. Comparative efficacy of novobiocin and amoxicillin in experimental sepsis caused by β-lactam-susceptible and highly resistant pneumococci. Int J Antimicrob Agents, 2010; 35(6):544–549; doi: 10.1016/j.ijantimicag.2010.02.007

75.

Enrique

, Di Ianni

, Goicoechea

, et al. New anticonvulsant candidates prevent P-glycoprotein (P-gp) overexpression in a pharmacoresistant seizure model in mice. Epilepsy Behav, 2021; 121(Pt B):106451; doi: 10.1016/j.yebeh.2019.106451

76.

Ahmad

, Abu-Taweel

, Aboshaiqah

, et al. The effects of quinacrine, proglumide, and pentoxifylline on seizure activity, cognitive deficit, and oxidative stress in rat lithium-pilocarpine model of status epilepticus. Oxid Med Cell Longev, 2014; 2014:630509; doi: 10.1155/2014/630509

77.

Ahmad

, Wadaan

, Pharm Sci

, et al. Ameliorating effects of proglumide on neurobehavioral and biochemical deficits in animal model of status epilepticus. Pak J Pharm Sci, 2014; 27(6):1945–1951

78.

Trinka

, Cock

, Hesdorffer

, et al. A definition and classification of status epilepticus—report of the ILAE task force on classification of status epilepticus. Epilepsia, 2015; 56(10):1515–1523; doi: 10.1111/epi.13121

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.20 MB

0.39 MB

5.75 MB

0.31 MB

0.02 MB