In this study, we isolated and characterized extracellular vesicles (EVs) from 37 TBI patients. The isolated EV population included EVs derived from neurons, and no significant differences in concentration or size were found between the placebo and biperiden groups, nor between patients who did or did not develop post-traumatic epilepsy (PTE).
Characterization of our sample revealed that 91.89% of the participants were male. According to DATASUS 2020 (the Brazilian healthcare system database), the majority of TBI cases in Brazil occur in males [1]. The most frequent cause of TBI in our sample, accounting for approximately 51.35% (n = 19), was traffic accidents. In the US, the most common cause is unintentional falls, followed by being unintentionally struck by an object and motor vehicle crashes [30], while Brazil's primary cause is traffic accidents [1]. These variations in causes of accidents highlight the influence of cultural factors on the etiology of TBI.
The clinical trial (NCT01048138), associated with the present study, was conducted to evaluate the efficacy of biperiden, a muscarinic anticholinergic drug, in reducing PTE. Biperiden had been previously reported to reduce hippocampal excitability and glutamate levels in a murine model with recurring epileptic seizures induced by pilocarpine. Biperiden treatment in this model also displayed reduced frequency and severity of spontaneous epileptic seizures, and decreased hippocampal hilus neuronal death [12]. Hilar cell loss is associated with temporal lobe epilepsy, both as a consequence as well as a potential contributor for this type of epilepsy [31, 32]. Temporal lobe lesions greatly increase the odds of developing PTE [33], and temporal lobe epilepsy is the most common type of PTE, followed by frontal lobe epilepsy. Less common types of PTE include parietal and occipital lobe epilepsy [34]. However, in this study, there was no significant difference regarding PTE development when comparing the number of individuals who took placebo and biperiden, even when stratifying for lesion location. No miRNA was found to be differentially expressed in TBI patients that developed PTE and those that did not, either.
Although we found no evidence of PTE reduction by biperiden treatment, we did observe a possible downregulation of EV miR-3615 in TBI patients that received biperiden. Considering our analysis thresholds, only 20 genes were found to be miR-3615 targets, which resulted in no pathway enrichments. These genes also had no variants associated with neither TBI nor epilepsy in previous GWAS. However, miR-3615 was previously found to be upregulated in CSF of pediatric severe TBI patients [35], indicating that it might be related to TBI response. Another EV miRNA, namely miR-9-5p, displayed reduced expression in biperiden-treated TBI patients.
No study had yet investigated miRNA changes due to biperiden treatment, and none of the miR-9-5p target genes were associated with any type of epilepsy, nor head injury, concussion, or nervous system injury in previous GWAS. However, biperiden treatment is often used for Parkinson’s disease (PD), and interestingly, miR-9-5p has been found upregulated in dopaminergic neurons of PD patients [36] and in a cell culture model of PD [37]. Upregulation of miR-9-5p has been found to be detrimental [37] and beneficial [38] to different PD models.
MiR-9-5p is a mature miRNA derived from miR-9 which has been extensively studied as a regulator of neurogenesis during neurodevelopment and in adult tissue [39], and has been used as a marker for exosomes derived from neurons found in blood serum [40]. Curiously, miR-9-5p is associated both with neuroprotection and neurodegeneration, having been found to be either up- or downregulated in different neurodegenerative diseases [39], and having several functions related to neuronal death and stress responses, including autophagy and hypoxia response [41], neuroinflammation [42, 43], and apoptosis [44]. It binds to and regulates mRNA of core regulatory genes for these processes, such as SIRT1 [37], [45] and REST [46], [44], which corroborates our enrichment analysis for genes regulated by this miRNA. Although no experimentally-validated gene present in miRTarBase regulated by miR-9-5p was found to be related to TBI nor PTE in GWAS catalog, several studies have linked this miRNA to either TBI or PTE.
Indeed, miR-9-5p has been suggested as a potential biomarker and as having a role in TBI and PTE [47]. MiR-9-5p upregulation occurs at 6 hours and 7 days after pilocarpine administration, a status epilepticus rat model, and overexpression of miR-9-5p decreased seizure latency and increased seizure intensity in this model [48]. However, seizure kindling in rat amygdala downregulates miR-9a-5p. This model also identified miR-9a-5p, the rat homolog of human miR-9-5p, as the only miRNA decreased in animals resistant to lamotrigine, an anti-epileptic agent [49], which suggests that while miR-9-5p levels are important to seizure formation, whether it is down- or upregulated depends on the stress signal. Interestingly, miR-9a-3p, derived from the same precursor miRNA as miR-9-5p, is upregulated in blood serum 2 days after fluid-percussion-induced TBI, though no miRNAs were found to predict PTE [50].
Regarding TBI, miR-9 has shown differential expression patterns: it was found to be downregulated in the rat cerebral cortex following TBI [51], but upregulated in the mouse ipsilateral hippocampus after TBI [52]. Additionally, increased plasma levels of miR-9a-3p at 2 days after experimental TBI distinguished injured rats from naive rats [53]. In a controlled cortical impact rat model of TBI, miR-9-5p expression in the traumatic foci is increased due to head trauma [42], peaking at 21 days after TBI. Further upregulation with a miR-9-5p mimetic, immediately after TBI, decreased neuronal death, neuroinflammation, and Modified Neurological Severity Score [42]. Conversely, downregulation of miR-9-5p, using an inhibitor 14 days after TBI, improved severity score and memory test, as well as promoted the proliferation of astrocytes and the release of astrocyte-derived neurotrophic factors, which promote synaptic remodeling and neurological function recovery [44]. These data suggest that miR-9-5p expression kinetics modulate TBI neurological outcomes.
Different studies have investigated the role of EVs miRNA in TBI, both in animal models and humans. In brain-derived EVs from blast injured mice, miR-9-5p was among the miRNA found to be upregulated, suggesting its potential association with TBI [21]. Human peripheral blood EVs also displayed an alteration in miR-9-5p levels in TBI patients, indiscriminately of TBI cause [21]. In contrast, miR-9 was found to be more abundant in EVs derived from cerebrospinal fluid (CSF) of non-injured patients compared to those with TBI [54]. Interestingly, miR-9-5p was not detected when profiling the whole serum of patients, regardless of TBI status [55]. This suggests that miR-9-5p may be exclusively present in EVs and possibly at low levels, as observed in our study. Indeed, it is highly expressed in brain tissue, but its expression in plasma, serum or blood samples is low (https://dianalab.e-ce.uth.gr/mited/#/expressions) [56], indicating its specific role in neuronal function and development. These findings highlight the distinct differential expression and distribution of miR-9 and miR-9-5p in various compartments (brain tissue, EVs, CSF, serum) and emphasize the complexity of miRNA regulation in the context of TBI and TBI resolution, which can be altered as a result of miR-9-5p expression levels.
Study limitations
This is a pioneer study that tests a potential drug for TBI-derived PTE, and searches for potential biomarkers that may also play a role in these conditions and their outcomes. Data reporting and methodology adhered to the recommendations from the International Society for Extracellular Vesicles [57]. However, our study has limitations that must be considered. Due to the small sample size, our results may not reflect a populational level. Our patients are challenging to recruit and maintain due to the highly invasive nature of their injuries. While miR-9-5p expression is generally considered to be mostly related to neuronal EVs, miRNA extraction was performed on serum EVs, which we found to display neuronal markers, but could also be secreted by other tissues. Indeed, the expression levels in our samples for this miRNA were low and to ensure the reliability of our findings, we performed multiple analyses using different pipelines. This rigorous approach was employed to enhance the confidence in our results. Additionally, our sample treatment and results are highly specific, making it difficult to establish direct correlations with existing literature. The BIPERIDEN clinical trial has been terminated, but a similar multicenter study [58] is now ongoing (NCT04945213) and our intention is to replicate the same methodology while also conducting additional analyses using a larger cohort of patients.
Furthermore, although some circulating and EV-bound miRNA have been previously proposed as biomarkers for some types of epilepsy, especially temporal lobe epilepsy and mesial temporal lobe epilepsy [59–61], no studies had yet investigated miRNA as biomarkers for PTE [62]. This is also the first study to associate biperiden treatment and miRNA expression, which can aid in elucidating its possible effects and mechanisms in PTE-associated TBI and other neurological conditions.