Skip to main content
NCBI home page
Cerebral Cortex (New York, NY) logoLink to Cerebral Cortex (New York, NY)
. 2022 Sep 20;33(7):4070–4084. doi: 10.1093/cercor/bhac327

N-acetylcysteine treatment mitigates loss of cortical parvalbumin-positive interneuron and perineuronal net integrity resulting from persistent oxidative stress in a rat TBI model

Mustafa Q Hameed 1,2,3, Nathaniel Hodgson 4, Henry H C Lee 5,6, Andres Pascual-Leone 7,8, Paul C MacMullin 9,10, Ali Jannati 11,12, Sameer C Dhamne 13,14, Takao K Hensch 15,16, Alexander Rotenberg 17,18,19,✉
PMCID: PMC10068300  PMID: 36130098

Abstract

Traumatic brain injury (TBI) increases cerebral reactive oxygen species production, which leads to continuing secondary neuronal injury after the initial insult. Cortical parvalbumin-positive interneurons (PVIs; neurons responsible for maintaining cortical inhibitory tone) are particularly vulnerable to oxidative stress and are thus disproportionately affected by TBI. Systemic N-acetylcysteine (NAC) treatment may restore cerebral glutathione equilibrium, thus preventing post-traumatic cortical PVI loss. We therefore tested whether weeks-long post-traumatic NAC treatment mitigates cortical oxidative stress, and whether such treatment preserves PVI counts and related markers of PVI integrity and prevents pathologic electroencephalographic (EEG) changes, 3 and 6 weeks after fluid percussion injury in rats. We find that moderate TBI results in persistent oxidative stress for at least 6 weeks after injury and leads to the loss of PVIs and the perineuronal net (PNN) that surrounds them as well as of per-cell parvalbumin expression. Prolonged post-TBI NAC treatment normalizes the cortical redox state, mitigates PVI and PNN loss, and - in surviving PVIs - increases per-cell parvalbumin expression. NAC treatment also preserves normal spectral EEG measures after TBI. We cautiously conclude that weeks-long NAC treatment after TBI may be a practical and well-tolerated treatment strategy to preserve cortical inhibitory tone post-TBI.

Keywords: traumatic brain injury, oxidative stress, parvalbumin-positive interneuron, perineuronal net, N-acetylcysteine

Introduction

Traumatic brain injury (TBI) is among the most common causes of morbidity and mortality globally (Faul and Coronado 2015; Dewan et al. 2018). Beyond the initial mechanical injury, TBI triggers molecular cascades that lead to increased and sustained oxidative stress (Bondy and LeBel 1993; Hsieh et al. 2017; Khatri et al. 2018; MacMullin et al. 2020) in the acute and subacute post-traumatic period, causing secondary injury and neuronal death (Ladak et al. 2019; Ng and Lee 2019). While symptoms such as headache and motor deficits may be evident immediately after TBI and typically resolve over time, continuing secondary damage leads to long-term sequelae such as post-traumatic epilepsy (PTE) and cognitive decline, often after a latent period of weeks to months, reflecting slow pathological processes (Wilson et al. 2017; Sharma et al. 2021).

Disruption of glutamate homeostasis is one mechanism that likely contributes to post-TBI oxidative stress. Glutamate spillage from dead and dying neurons and decreased synaptic glutamate clearance due to reduced astrocytic glutamate transporter expression (Goodrich et al. 2013; Hameed et al. 2019) and function (Dorsett et al. 2017) results in persistently elevated extracellular glutamate levels (Yamamoto et al. 1999; Zhuang et al. 2019). The subsequent increased post-traumatic excitatory tone increases the metabolic demand and production of reactive oxygen species (ROS; Bondy and LeBel 1993; Khatri et al. 2018). In parallel, extracellular glutamate accumulation can also cause oxidative toxicity independent of ionotropic receptor overactivation by interfering with the system xc− cystine/glutamate antiporter (Sxc−) responsible for cystine uptake for the generation of glutathione (GSH), the most abundant endogenous intracellular antioxidant (Murphy et al. 1989).

Parvalbumin-positive interneurons (PVIs), representing close to 40% of the cortical gamma-aminobutyric acid (GABA)ergic inhibitory interneuron population (Rudy et al. 2011), are particularly vulnerable to post-traumatic excitotoxic and oxidative stress (Moga et al. 2002; Cabungcal et al. 2013a; Ruden et al. 2021). PVIs have a high baseline firing rate and therefore produce high ROS concentrations which can be skewed toward toxic when glutamatergic tone is increased, as happens after TBI (Yamamoto et al. 1999; Cantu et al. 2015; Zhuang et al. 2019). Increased ambient ROS also degrade the perineuronal net (PNN) which is critical for PVI survival (Cabungcal et al. 2013b; Wen et al. 2018), further endangering these interneurons. Given their prominent role in maintaining cortical excitation:inhibition (E:I) balance (Sohal et al. 2009; Cammarota et al. 2013), the biased loss of this vulnerable cell population (Hsieh et al. 2017) may be a large contributor to PTE and neurocognitive decline (Godoy et al. 2022).

N-acetylcysteine (NAC), a well-tolerated and inexpensive drug, has been used to supplement intracellular cysteine and ultimately reduce oxidative cell injury for decades (Tenorio et al. 2021). Plausibly (as tested in the present experiments), acute post-TBI treatment with NAC may mitigate injury to neuronal elements that are vulnerable to oxidative stress, such as PVIs. Indeed, NAC treatment has been tested extensively in TBI models (Ghiam et al. 2021). Yet, NAC utility comes with an important caveat relevant to post-TBI pathophysiology: NAC deacetylation to cysteine, oxidation to cystine, and subsequent transport via Sxc− in exchange for glutamate may lead to unwanted, further increases in ambient glutamate after injury (Reissner 2014; Soria et al. 2014; Wright et al. 2016). Therefore, the efficacy of NAC treatment in normalizing cortical redox reserve and mitigating PVI and PNN dropout post-TBI requires explicit testing. In complement to our earlier work showing increased oxidative stress and PVI and PNN loss following TBI (Hsieh et al. 2017; MacMullin et al. 2020), we now test, in a rat TBI model, whether NAC treatment improves cortical redox equilibrium as measured by the ratio of reduced glutathione (GSH) to oxidized cortical glutathione disulfide (GSSG), PVI number and health, PNN integrity, and electroencephalographic (EEG) measures of injury after TBI.

Materials and methods

Animals

Twelve week old male Sprague–Dawley rats (367 ± 36 g) were housed in standard cages with ad libitum supply of food and water and a 12 h light/dark cycle. All procedures were approved by the Institutional Animal Care and Use Committee at Boston Children’s Hospital and were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All efforts were made to minimize the number of rats used in the present study.

Experimental groups

Rats were randomly divided into 3 experimental groups as follows: (i) Sham TBI; (ii) TBIH2O, intraperitoneal (IP) normal saline once daily on post-operative day (POD) 0 and 1, regular drinking water from POD 3 onward; (iii) TBINAC, 100 mg/kg IP NAC once daily on POD 0 and 1, 1 g/L NAC per-oral (PO) in drinking water (corresponding to approximately 100 mg/kg/day) from POD 3 onward.

Electroencephalography

One cohort of 3 experimental groups (Sham, TBIH2O, and TBINAC) was generated to obtain serial video-EEG 3 and 6 weeks after sham or verum injury (N = 5–7/group; Fig. 1A).

Fig. 1.

Fig. 1

Open in a new tab

Study design. A) Experimental timeline and cohorts. Rats were exposed to FPI or Sham injury in cohorts of rats at week 0. Verum injured rats were further assigned to vehicle control (TBIH2O) or NAC (TBINAC) treatment arms. Brain tissue from individual cohorts was harvested for HPLC or IHC at 3 and 6 weeks after injury. In a separate cohort (Sham, TBIH2O, and TBINAC), EEG implant surgery was performed at week 0, immediately following injury. Serial video-EEG was recorded for 48 h at 3 and 6 weeks. B) Schematic dorsal view of the rat skull showing the area over the left convexity where FPI was induced (solid black circle), and epidural EEG electrode implantation sites over the right olfactory bulb (reference) and left cortex posterior to the lesion (active). C) Schematic, en face coronal view of a rat brain delineating the FPI-induced lesion in the left parietal cortex (shown on the right in this view). HPLC and IHC analysis was performed in perilesional cortex (PL) medial to the lesion and in the corresponding contralesional cortex (CL).

High-performance liquid chromatography and immunohistochemistry

Four cohorts of 3 groups (Sham, TBIH2O, and TBINAC) each were generated and allocated to quantify redox reserve by high-performance liquid chromatography (HPLC) and histological measures via immunohistochemistry (IHC). Cortical tissue was harvested 3 and 6 weeks after Sham or verum injury (NHPLC = 5–6/group/timepoint; NIHC = 6–8/group/timepoint; Fig. 1A).

Fluid percussion injury

TBI was induced as previously described (Hameed et al. 2019). Briefly, rats were anesthetized using isoflurane vapor and mounted on a stereotactic frame (Stoelting Co., IL, United States). Body temperature was maintained using a circulating water heating pad set to 37 °C. A 4 mm diameter craniotomy was centered over the left sensorimotor cortex, 2 mm posterior to the bregma, with the lateral edge of the craniotomy adjacent to the lateral ridge (Fig. 1B and C), and the dura was examined to confirm its integrity. A length of plastic tubing attached to a modified pipette tip with a 4 mm aperture at its free end was fitted to the male luer connector of the fluid percussion device (AmScien Instruments, VA, United States), and the connection was made airtight using polytetrafluoroethylene tape. The plastic tubing was filled with sterile saline at room temperature and was inspected to confirm the absence of air bubbles. The tip was positioned over the exposed dura and was held in place using the micromanipulator arm of the frame, with the tip edge in tight contact with the skull. A percussion wave of 2.3 ± 0.1 atm was delivered to induce moderate TBI. Each rat’s respiratory rate was visually monitored throughout the TBI procedure, and epochs of apnea after injury were recorded. Rats that did not resume spontaneous breathing within 30 s of fluid percussion injury (FPI) were excluded from the study to standardize injury severity. Sham control rats underwent all surgical procedures without FPI. Opioid analgesia (sustained-release buprenorphine; 1.2 mg/kg) was administered subcutaneously for 72 h of post-operative coverage.

E‌EG electrode implantation and acquisition

Rats were implanted with wireless telemetry transmitters (PhysioTel ETA-F10; Data Sciences International, MN, United States) with the active electrode over ipsilesional cortex 2 mm posterior to the FPI craniotomy, and reference electrode over the contralesional olfactory bulb (Fig. 1B; Goodrich et al. 2013), to record 1-channel continuous video-EEG (sampling rate: 1,000 Hz), core-body temperature (0.1 Hz), and home-cage locomotion activity (0.1 Hz; Dhamne et al. 2017; Kelly et al. 2018). Data were collected for 48 h per timepoint; the first 24 h corresponded to an acclimation period and only the last 24 h were analyzed.

Measurement of GSH and GSSG

Sample preparation

Rats were decapitated using a sharp guillotine, and fresh perilesional and contralesional tissue was dissected (Fig. 1C) and snap-frozen before being processed as described earlier (MacMullin et al. 2020). Briefly, frozen tissue was homogenized (10% weight/volume) in phosphate-buffered saline (PBS; pH 5.5) and sonicated for 15 s on ice. A 20 μL aliquot was removed for protein quantification. The remaining sonicate was added 10:1 to 0.4 N perchloric acid and centrifuged at 13,000 rpm for 60 min at 4 °C. All samples were prepared at 4 °C and with a pH well below the 8.8 pKa of the glutathione thiol group. GSH is much more stable under these sample preparation conditions, minimizing artefactual oxidation to GSSG.

HPLC

Samples were analyzed as described previously (MacMullin et al. 2020). Our HPLC system consisted of a Beckman System Gold auto sampler and pump (Beckman Coulter, CA, United States) a ZORBAX Eclipse XDB-C8 column (3 × 150 mm, 3.5 μm; Agilent Technologies Inc., CA, United States), and a CouloChem III electrochemical detector (Thermo Fisher Scientific, MA, United States) with a boron-doped diamond electrode (Model 5040; Thermo Fisher Scientific) run at a potential of 1,500 mV. Analytes were resolved using a dual mobile phase gradient elution. Both mobile phases A and B contained 25 mM of sodium phosphate adjusted to pH 2.65 with phosphoric acid and 2.1 mM 1-octanesulfonic acid as an ion pairing agent. Mobile phase B (MP-B) contained 50% acetonitrile. The gradient ran 0% MP-B from 0 to 8 min and then ramped to 30% MP-B from 8 to 20 min. The system returned to 0% MP-B from 25 to 36 min.

Analytes were quantified based on standard curves for each compound (32 Karat software v8.0; Beckman Coulter, CA, United States). Samples were normalized to total protein content. GSH and GSSG levels and GSH/GSSG ratios in injured rats were normalized to Sham controls for analysis.

Histology

Perfusion of cortical tissues

Deeply anesthetized rats were transcardially perfused with PBS followed by 4% paraformaldehyde (PFA), as previously described (MacMullin et al. 2020). Harvested brains were then immersion-fixed in 4% PFA at 4 °C for 24 h before being cryopreserved in 30% sucrose. Cryoprotected brains were rapidly frozen in Tissue-Plus OCT Compound (Thermo Fisher Scientific, MA, United States) and were stored at −80 °C for at least 48 h before sectioning.

IHC

Free-floating coronal sections (30 μm; Bregma −2 mm; Fig. 1C) were blocked with 10% goat serum and 0.1% Triton X-100 in PBS for 1 h and then incubated with primary antibodies overnight at 4 °C (PVI: antiparvalbumin antibody; Swant, Marly, Switzerland; PNN: biotinylated wisteria floribunda agglutinin [WFA]; MilliporeSigma, MO, United States) followed by Alexa Fluor conjugated secondary antibodies for 1 h at room temperature (PVI: antiguinea pig Alexa Fluor 488; PNN: streptavidin Alexa Fluor 594; Thermo Fisher Scientific, MA, United States; MacMullin et al. 2020). Stained sections were mounted using Fluoromount medium containing 4′,6′-diamidino-2-phenylindole (DAPI) nuclear counterstain (Southern Biotechnology, Birmingham, AL). All processing and staining steps were performed under identical conditions using the same batch of buffers to minimize intersample variability.

Image acquisition

Perilesional sites in the parietal cortex (100 μm medial from the medial border of the lesion) and corresponding contralesional sites (Fig. 1C) were identified as described previously (Hsieh et al. 2017). 1,024 × 1,024 μm confocal images were acquired at 10× magnification using FV10-ASW software (v2.1C) on an FV1000 confocal laser scanning microscope (Olympus Corporation, Tokyo, Japan) with the following parameters: Channel 1: 488 nm laser power = 2%, Dye = Alexa Fluor 488, PMT voltage = 465, PMT gain = 1; Channel 2: 561 nm laser power = 5%, Dye = Alexa Fluor 594, PMT voltage = 585, PMT gain = 1. Individual channels were acquired sequentially and at least 4 sections were imaged per rat.

Cell count

Images were analyzed using ImageJ software (Hsieh et al. 2017; MacMullin et al. 2020). First, a 400 μm wide region of interest corresponding to layers I–VI was selected in the middle of the imaging field. Identical threshold settings and filters for size (50–5,000 μm2) and circularity (0–1.0) were then applied to all images. Structures positively stained for PV and/or WFA (PNN) were identified and visually confirmed to be cells using DAPI nuclear counterstaining, and automated counts were then recorded. Results from TBIH2O and TBINAC rats were normalized to Sham controls per hemisphere and timepoint.

PV intensity

To quantify PV content in individually identified PNN+ cells, images were overlaid with a PV immunofluorescence channel using ImageJ, and intensity per cell was measured. PV intensity data were normalized to Sham controls for analysis.

E‌EG spectral analysis

Raw time-series EEG data were preprocessed using a custom algorithm (MATLAB R2019b; MathWorks Inc., Natick MA) employing a combination of amplitude thresholding and time windowing to remove movement-related artifacts and electrical noise. EEG data were divided into nonoverlapping 30-s blocks, and a continuous wavelet transform was performed using analytic Morse wavelets to obtain a time-frequency decomposition for each block. The time-decomposed data were further segmented into “quiet” epochs only during the 12 h light period (7 am–7 pm), when locomotion activity measured 0, to minimize motion artifacts. The processed decomposed EEG signal in the “quiet” epochs was used to compute absolute spectral power binned at 0.5 Hz, which was then normalized to total EEG power (0.5–90 Hz) to account for intersubject variability (MacMullin et al. 2020). Total spectral power in the delta (0.5–4 Hz), theta (4–13 Hz), and beta (13–30 Hz) frequency bands was compared between groups at each timepoint.

Further, time-series EEG data were also analyzed for epileptic seizures and spikes using a semiautomated seizure detection algorithm (Neuroscore 3.4.1; Data Sciences International, MN, United States). Automatically marked events were validated by manual review of video-EEG (Kelly et al. 2018; Purtell et al. 2018).

Statistical analysis

Data were compared across experimental groups at each timepoint using one-way ANOVA with post- hoc Tukey’s multiple comparisons tests (Prism 9; GraphPad Software Inc, CA, United States) with the significance level set at P < 0.05. Given that per-rat data were averaged, and then individual means were averaged to obtain group data, all results are presented as mean ± standard error of the mean (SEM).

Results

NAC treatment attenuates cortical oxidative stress after TBI

Change in redox state is both an indicator of injury, and pathogenic. Relevant to this work, oxidative stress can potentiate excitotoxicity and can decrease PV and PNN expression. The ratio of GSH to GSSG indicates the redox state (Schafer and Buettner 2001), with higher ratios denoting a more reducing environment and lower ratios being more oxidizing.

TBI resulted in significant decreases in the perilesional GSH/GSSG ratio in TBIH2O rats compared to Sham controls both 3 and 6 weeks after injury (3 W: −29.88% ± 8.95%, P = 0.037; 6 W: −26.43% ± 7.83%, P = 0.004), with NAC treatment normalizing the ratio in TBINAC rats compared to Sham at both timepoints (3 W: −14.49% ± 12.70%, n.s.; 6 W: −7.26% ± 3.26%, n.s.; Fig. 2A and B). Six weeks after injury, perilesional GSH/GSSG was also significantly lower in TBIH2O compared to TBINAC rats (P = 0.037; Fig. 2B).

Fig. 2.

Fig. 2

Open in a new tab

Intracellular redox status in perilesional and contralesional cortex. Cortical GSH/GSSG ratio as measured using HPLC. Note: A lower ratio indicates increased oxidative stress. A) TBI resulted in a decrease in perilesional but not contralesional GSH/GSSG, with NAC treatment attenuating this drop, 3 weeks after injury. B) Six weeks after injury, perilesional GSH/GSSG was depressed in TBIH2O rats compared to Sham as well as compared to TBINAC rats. There was no difference in GSH/GSSG between Sham and TBINAC rats in either hemisphere. C, D) The post-traumatic decrease in perilesional GSH/GSSG was driven primarily by an increase in GSSG in TBIH2O rats compared to Sham controls. GSSG levels did not change in TBINAC rats compared to Sham and were significantly lower than in TBIH2O animals bilaterally 6 weeks after TBI. E, F) While GSH did not change in TBIH2O rats, levels decreased in TBINAC animals bilaterally compared to Sham 6 weeks after TBI. (Data shown as mean ± SEM; *P < 0.05, **P < 0.01).

The post-traumatic decrease in perilesional GSH/GSSG ratio was driven primarily by an increase in GSSG in TBIH2O rats compared to Sham controls (3 W: +43.22% ± 16.80%, n.s. [P = 0.052]; 6 W: +33.07% ± 14.92%, P = 0.046). On the other hand, GSSG levels did not change in TBINAC rats compared to Sham (3 W: +12.44% ± 14.89%, n.s.; 6 W: −5.95% ± 5.08%, n.s.) and were significantly lower than in TBIH2O animals 6 weeks after TBI (P = 0.024; Fig. 2C and D). Interestingly, while perilesional GSH did not change in TBIH2O rats (3 W: −5.09% ± 2.72%, n.s.; 6 W: −6.57% ± 1.81%, n.s.), levels decreased in TBINAC rats compared to Sham 6 weeks after TBI (3 W: −9.29% ± 4.60%, n.s.; 6 W: −12.74% ± 3.15%, P = 0.009; Fig. 2E and F).

There was no effect of injury or treatment on contralesional GSH/GSSG ratios 3 weeks (TBIH2O: −6.09% ± 3.96%, n.s.; TBINAC: −7.66% ± 2.38%, n.s.) or 6 weeks after TBI (TBIH2O: −9.63% ± 4.22%, n.s. [P = 0.057]; TBINAC: −5.09% ± 2.10%, n.s.; Fig. 2A and B). GSSG and GSH levels did not differ between groups 3 weeks post-injury (GSSG − TBIH2O: +4.57% ± 2.31%, TBINAC: +5.00% ± 3.00%; GSH − TBIH2O: −2.01% ± 2.12%, TBINAC: −2.96% ± 1.69%; n.s.; Fig. 2C and E). GSSG and GSH levels in TBIH2O rats remained unchanged compared to Sham 6 weeks post-TBI (GSSG: +6.07% ± 3.85%; GSH: −4.57% ± 2.06%; n.s.); however, TBINAC rats exhibited a significant decrease in GSSG (−8.50% ± 2.50%) compared to TBIH2O (P = 0.021) and a decrease in GSH (−13.08% ± 2.19%) compared to Sham controls (P = 0.003; Fig. 2D and F).

NAC treatment attenuates cortical PVI loss

A dramatic reduction of PVI cell count was observed in the perilesional cortex 3 weeks after injury in TBIH2O rats compared to Sham and TBINAC animals (Sham: 100.00% ± 5.43%; TBIH2O: 64.98% ± 4.33%; TBINAC: 116.80% ± 4.27%; Sham vs. TBIH2O: P < 0.001; TBIH2O vs. TBINAC: P < 0.0001; Fig. 3A–D). PVI counts remained significantly lower in untreated, injured TBIH2O rats 6 weeks after injury compared to Sham and TBINAC animals (Sham: 100.00% ± 8.01%, TBIH2O: 58.29% ± 5.39%; TBINAC: 99.08% ± 5.36%; Sham vs. TBIH2O: P < 0.001; TBIH2O vs. TBINAC: P < 0.001; Fig. 3H–K).

Fig. 3.

Fig. 3

Open in a new tab

Parvalbumin-positive interneurons in perilesional and contralesional cortex. A) PVI quantity decreased in both perilesional and contralesional cortex 3 weeks after injury and was rescued with NAC treatment following injury. B–G) Representative confocal images showing the staining of PVI in perilesional and contralesional cortex in all experimental groups (Sham, TBIH2O, and TBINAC) 3 weeks after injury. H) PVI count remained depressed bilaterally 6 weeks after injury, with NAC treatment attenuating PVI loss. I–N) Representative images showing perilesional and contralesional PVI staining 6 weeks after injury. (Data shown as mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Similarly, contralesional PVI count was also reduced in TBIH2O rats compared to Sham and TBINAC rats 3 weeks (Sham: 100.00% ± 5.29%; TBIH2O: 62.42% ± 8.41%; TBINAC: 113.10% ± 2.99%; Sham vs. TBIH2O: P = 0.002; TBIH2O vs. TBINAC: P < 0.001; Fig. 3A and E–G) and 6 weeks after injury (Sham: 100.00% ± 9.09%; TBIH2O: 64.55% ± 4.84%; TBINAC: 105.50% ± 11.35%; Sham vs. TBIH2O: P = 0.029; TBIH2O vs. TBINAC: P = 0.009; Fig. 3B and L–N).

NAC treatment prevents progressive post-traumatic cortical PNN degradation

Immunohistochemical analysis of PNN+ cells in perilesional cortex revealed significant PNN loss following TBI, with NAC treatment mitigating this decline 3 weeks (Sham: 100.00% ± 2.07%; TBIH2O: 84.43% ± 4.43%; TBINAC: 98.26% ± 2.93%; Sham vs. TBIH2O: P = 0.010; TBIH2O vs TBINAC: P = 0.026; Fig. 4A–D) and 6 weeks after injury (Sham: 100.00% ± 4.32%; TBIH2O: 72.02% ± 5.09%; TBINAC: 94.47%± 4.27%; Sham vs. TBIH2O: P = 0.001; TBIH2O vs TBINAC: P = 0.006; Fig. 4H–K).

Fig. 4.

Fig. 4

Open in a new tab

PNN count in perilesional and contralesional cortex. A) PNN+ cell counts decreased bilaterally 3 weeks after TBI and were salvaged with NAC treatment following injury. B–G) Representative confocal images showing PNN staining in perilesional and contralesional cortex in all experimental groups (Sham, TBIH2O, and TBINAC) 3 weeks after injury. H) Perilesional PNN+ cell counts remained depressed 6 weeks after injury compared to Sham and were rescued by NAC treatment; however, there was no difference in contralesional counts between TBIH2O and TBINAC groups at this timepoint. I–N) Representative images showing perilesional and contralesional PVI staining 6 weeks after injury. (Data shown as mean ± SEM; *P < 0.05, **P < 0.01).

TBI also affected PNN+ cell counts in the contralesional cortex, with NAC again rescuing this loss both 3 weeks (Sham: 100.00% ± 4.58%; TBIH2O: 81.23% ± 2.29%; TBINAC: 105.20% ± 4.42%; Sham vs TBIH2O: P = 0.014; TBIH2O vs TBINAC: P = 0.004; Fig. 4A and E–G) and 6 weeks after injury (Sham: 100.00% ± 4.11%; TBIH2O: 82.01% ± 3.97%; TBINAC: 89.76% ± 2.14%; Sham vs TBIH2O: P = 0.005; Fig. 4H and L–N).

NAC treatment prevents post-traumatic loss of per-cell PV content

PV content in PNN+ cells was further analyzed by co-immunofluorescence of PV and PNN. TBI caused a leftward shift in the mode of the frequency distribution of PV intensity per identified PNN+ cell bilaterally in both TBIH2O (bin center: perilesional, 400 arbitrary units [AU]; contralesional, 400 AU) and TBINAC rats (perilesional, 400 AU; contralesional, 400 AU) compared to Sham (perilesional, 600 AU; contralesional, 600 AU; Fig. 5A) 3 weeks after injury. While there was no significant difference in the percentage of total PNN+ cells with low PV expression (i.e. < 100 AU of PV content) at this timepoint in either the perilesional (Sham: 0.29% ± 0.23%; TBIH2O: 0.73% ± 0.29%; TBINAC: 0.19% ± 0.08%; n.s.) or contralesional cortex (Sham: 0.11% ± 0.11%; TBIH2O: 0.08% ± 0.08%; TBINAC: 0.08% ± 0.08%; n.s.; Fig. 5B), TBINAC rats exhibited a significantly higher percentage of total cortical PNN+ cells expressing high levels (>1,500 AU) of per-cell PV in both perilesional (Sham: 3.56% ± 1.07%; TBIH2O: 3.67% ± 1.12%; TBINAC: 24.12% ± 2.12%; Sham vs. TBINAC: P < 0.0001; TBIH2O vs. TBINAC: P < 0.0001) and contralesional cortex (Sham: 3.81% ± 0.66%; TBIH2O: 3.03% ± 0.98%; TBINAC: 26.18% ± 4.13%; Sham vs. TBINAC: P < 0.0001; TBIH2O vs. TBINAC: P < 0.0001; Fig. 5C).

Fig. 5.

Fig. 5

Open in a new tab

PV expression in PNN+ cells. A) Frequency distribution of PV pixel intensity of PNN+ cells in perilesional and contralesional cortex 3 weeks after injury in all experimental groups (Sham, TBIH2O, and TBINAC). PV pixel intensity is binned at 200 AU; x-axis shows bin centers. B) There was no difference between groups in the relative frequency of PNN+ cells with low PV expression (<100 AU of pixel intensity) in perilesional and contralesional cortex 3 weeks after injury. C) Three weeks after injury, there were significantly higher proportions of PNN+ cells with high PV expression (>1,500 AU of pixel intensity) bilaterally in TBINAC rats compared to both Sham and TBIH2O animals. D) Summary plots of PV pixel intensity in PNN+ cells in perilesional and contralesional cortex 6 weeks after injury. E) There was a significant increase bilaterally in PNN+ cells with low PV in TBIH2O rats compared to both Sham and TBINAC rats 6 weeks after injury. F) NAC treatment induced high PV expression in a significantly higher proportion of perilesional PNN+ cells 6 weeks after injury compared to Sham and TBIH2O rats. (Data shown as mean ± SEM; *P < 0.05, **P < 0.01, ****P < 0.0001).

Six weeks after injury, TBIH2O rats continued to exhibit a leftward modal shift in the frequency distribution of PV intensity per identified PNN+ cell in both hemispheres (perilesional, 400 AU; contralesional, 200 AU) compared to both Sham (perilesional, 600 AU; contralesional, 800 AU) and TBINAC rats (perilesional, 800 AU; contralesional, 800 AU; Fig. 5D). In addition, TBIH2O rats demonstrated a significantly higher percentage of total PNN+ cells with low PV expression at this timepoint in both the perilesional (Sham: 1.03% ± 0.50%; TBIH2O: 9.84% ± 3.21%; TBINAC: 0.00% ± 0.00%; Sham vs. TBIH2O: P = 0.029; TBINAC vs. TBIH2O: P = 0.011) and contralesional cortex (Sham: 1.40% ± 1.24%; TBIH2O: 7.09% ± 1.22%; TBINAC: 0.32% ± 0.32%; Sham vs. TBIH2O: P = 0.006; TBINAC vs. TBIH2O: P = 0.001; Fig. 5E). TBINAC rats continued to exhibit significantly higher percentages of PNN+ cells with high PV content in perilesional (Sham: 1.90% ± 0.53%; TBIH2O: 2.24% ± 1.40%; TBINAC: 14.57% ± 4.20%; Sham vs. TBINAC: P = 0.023; TBIH2O vs. TBINAC: P = 0.019) but not in contralesional cortex (Sham: 8.80% ± 5.96%; TBIH2O: 4.48% ± 2.26%; TBINAC: 15.18% ± 6.11%; n.s.; Fig. 5F).

NAC treatment mitigates progressive post-traumatic changes in EEG power

Spectral analysis of 12 h light period (7 am–7 pm) EEG data (Fig. 6A and B) revealed a significant increase in relative power in the delta (δ; 0.5–4 Hz) EEG band in TBIH2O rats compared to Sham controls both 3 weeks (Sham: 100.00% ± 3.52%; TBIH2O: 119.70% ± 7.01%; TBINAC: 98.57% ± 3.53%; Sham vs. TBIH2O: P = 0.044; TBIH2O vs TBINAC: P = 0.019; Fig. 6C and D) and 6 weeks (Sham: 100.00% ± 3.20%; TBIH2O: 120.30% ± 6.66%; TBINAC: 96.42% ± 3.90%; Sham vs. TBIH2O: P = 0.035; TBIH2O vs TBINAC: P = 0.008) following TBI (Fig. 6E and F), with NAC treatment renormalizing this rise at both timepoints.

Fig. 6.

Fig. 6

Open in a new tab

EEG spectral analysis. Relative EEG power spectra in all experimental groups (Sham, TBIH2O, and TBINAC) A) 3 weeks and B) 6 weeks after Sham or verum injury. C) Callout showing relative power across the delta (δ) frequency range (0.5–4 Hz) in all experimental groups 3 weeks after injury. D) There was a significant increase in total delta power in TBIH2O rats compared to Sham controls and TBINAC rats 3 weeks post-TBI. E) Relative power across the delta range 6 weeks after injury. F) Six weeks after TBI, total delta power remained significantly increased in TBIH2O rats. G) Relative power across the theta (θ) range (4–13 Hz) 3 weeks after TBI. H) Three weeks post-TBI, total theta power trended lower in TBIH2O rats compared to Sham and was significantly lower compared to TBINAC rats. I) Relative power across the theta range 6 weeks after TBI. J) Six weeks post-injury, total theta band power was significantly lower in TBIH2O rats compared to both Sham and TBINAC animals. K) Relative power in the beta (β) range (13–30 Hz) 3 weeks after injury. L) Total beta power was significantly lower in TBIH2O rats compared to Sham and TBINAC animals 3 weeks post-TBI. M) Relative power across the beta range 6 weeks after injury. N) Total beta power remained significantly lower in TBIH2O rats 6 weeks after TBI. (Data shown as mean ± SEM; *P < 0.05, **P < 0.01).

Conversely, relative power in the theta (θ; 4–13 Hz) EEG band trended lower in TBIH2O rats compared to Sham and was significantly lower compared to TBINAC rats 3 weeks after injury (Sham: 100.00% ± 3.23%; TBIH2O: 84.30% ± 6.04%; TBINAC: 103.70% ± 2.39%; Sham vs. TBIH2O: P = 0.059; TBIH2O vs. TBINAC: P = 0.014; Fig. 6G and H). Six weeks after TBI, the total θ band power was significantly lower in TBIH2O rats compared to both Sham and TBINAC animals (Sham: 100.00% ± 2.72%; TBIH2O: 84.01% ± 5.67%; TBINAC: 104.50 ± 3.13; Sham vs. TBIH2O: P = 0.048; TBIH2O vs. TBINAC: P = 0.009; Fig. 6I and J).

TBIH2O rats also exhibited significant decreases in the total beta (β) band power compared to Sham and TBINAC animals both 3 weeks (Sham: 100.00% ± 4.34%; TBIH2O: 76.78% ± 6.26%; TBINAC: 99.31% ± 3.66%; Sham vs. TBIH2O: P = 0.015; TBIH2O vs. TBINAC: P = 0.010; Fig. 6K and L) and 6 weeks after injury (Sham: 100.00% ± 5.42%; TBIH2O: 78.02% ± 6.33%; TBINAC: 108.10% ± 5.01%; Sham vs. TBIH2O: P = 0.048; TBIH2O vs. TBINAC: P = 0.004; Fig. 6M and N). There was no significant difference between Sham and TBINAC rats in any EEG power band at any timepoint after injury (Fig. 6C–N).

No electrographic seizures were detected in any rat, and subepileptic spike counts were similar across all cohorts 3 weeks (Sham: 100.00% ± 54.98%; TBIH2O: 233.33% ± 97.66%; TBINAC: 107.41% ± 38.96%; n.s.) and 6 weeks after TBI (Sham: 100.00% ± 50.89%; TBIH2O: 246.43% ± 88.21%; TBINAC: 103.57% ±52.16%; n.s.). Furthermore, analysis of 12 h light (7 am–7 pm), 12 h dark (7 pm–7 am), and 24 h (7 am–7 am) activity data did not reveal any differences between the groups in total inactive time (presumed sleep) or inactive–active transitions 3 or 6 weeks post-TBI to indicate sleep disturbances that might confound spectral analysis (Fig. S1).

Discussion

We previously demonstrated that pathologic increases in oxidative stress (Hsieh et al. 2017) and progressive PVI injury (Hsieh et al. 2017; MacMullin et al. 2020) are components of post-TBI pathophysiology which may be targeted by therapeutic interventions (Hameed et al. 2019). Loss of cortical inhibitory tone weeks after TBI might underlie chronic TBI sequelae such as PTE and cognitive dysfunction (Hsieh et al. 2017; Wilson et al. 2017; MacMullin et al. 2020), and preclinical studies indicate that reestablishing redox equilibrium early after the primary insult may limit secondary neuronal injury (Cabungcal et al. 2013a; Kochanek et al. 2015). NAC has been used to reduce oxidative stress and improve recovery in patients after TBI, albeit only for the short-term and with varying degrees of success (Bhatti et al. 2017; Ghiam et al. 2021). We now report for the first time that the decrease in the cortical GSH/GSSG ratio (which indicates oxidative stress) persists for up to 6 weeks after moderate TBI. In addition, we address concerns discussed earlier regarding NAC’s excitotoxic potential following TBI by showing that long-term NAC treatment after injury attenuates (rather than exacerbates) oxidative stress, mitigates PVI and PNN loss, and preserves spectral EEG measures after TBI.

TBI results in persistent cortical oxidative stress, which is reversed by NAC treatment

We previously reported that 8-hydroxy-2′-deoxyguanosine (8-oxo-dG - a major product of oxidative stress-mediated deoxyribonucleic acid damage) levels peak 4 weeks after TBI (Hsieh et al. 2017). We now identify a potential causative factor for this increase in an oxidative stress biomarker and show that TBI-induced redox imbalance, as indexed by the GSH/GSSG ratio (Schafer and Buettner 2001; Zitka et al. 2012) peaks early after TBI and plateaus thereafter until at least 6 weeks injury.

Given that post-TBI excitotoxicity generates excess ROS (Khatri et al. 2018; Zhuang et al. 2019), our data indicate that increased GSH consumption and conversion to GSSG is the primary cause of post-traumatic glutathione imbalance. Regional GSH and GSSG abundance is also in part contingent on glutathione reductase (GR), which reduces GSSG to the active form GSH (Lu 2009, 2013). Relevant to our findings, the GR capacity to reduce GSSG to GSH can be overwhelmed by persistent oxidative stress (Lu 2009, 2013), thus further depleting intracellular GSH and reducing the GSH/GSSG ratio as excess GSSG is either shuttled extracellularly or conjugated to proteins to stave off major shifts in redox equilibrium (Lu 2009, 2013).

Sxc− also plays an important role in the regulation of the intracellular GSH/GSSG redox couple (Lewerenz et al. 2013) by stoichiometrically exchanging extracellular cystine (the oxidized form of cysteine) for intracellular glutamate and providing a source of cysteine for GSH synthesis. Excess extracellular glutamate in the TBI setting (Zhuang et al. 2019) may also act as a competitive inhibitor for cystine uptake via Sxc−, resulting in further intra- and extracellular antioxidant dysregulation (Lewerenz et al. 2013; Dixon et al. 2014). Of note, our data show a trend toward a decrease in GSH levels in TBIH2O rats.

The aforementioned persistent oxidative stress presumably contributes to secondary neuronal injury (Khatri et al. 2018), particularly PVI and PNN damage, as reported previously (Hsieh et al. 2017; MacMullin et al. 2020) and reproduced in the present work. Our data indicate that exogenous NAC administration prevents TBI-induced decreases in perilesional GSH/GSSG ratios and attenuates intermediate histological and neurophysiological sequelae. While NAC has low blood–brain barrier penetrance, it is deacetylated and oxidized in the liver and increases serum and tissue levels of cystine. Astrocytes take up cystine through the Sxc− and reduce it to cysteine, which is ultimately shuttled to neurons which lack a cystine transporter. Neurons may then utilize cysteine to synthesize intracellular GSH (Aoyama et al. 2008). Interestingly, with NAC treatment, we detect a decrease rather than an increase in both perilesional and contralesional GSH in TBINAC rats in the presence of a normal GSH/GSSG ratio. While this decrease appears paradoxical when viewed in isolation, it is perhaps less surprising when a systems biology approach is considered in addition to the fact that a tissue’s redox potential is defined by the GSH/GSSG ratio rather than GSH or GSSG alone (Schafer and Buettner 2001). Basal expression of glutamate cysteine ligase (GCL), the rate-limiting enzyme in GSH synthesis, is regulated by products of lipid peroxidation (Lu 2009, 2013), and we propose that a NAC-mediated increase in cysteine - a potent antioxidant in its own right - decreases overall intracellular oxidative stress independent of GSH (Patriarca et al. 2005; Ezerina et al. 2018). GCL is also under negative feedback inhibition by GSH, and perhaps higher intracellular cysteine levels available after NAC treatment transiently increase de novo GSH synthesis. Either scenario, or both, may reduce oxidative stress below physiological levels and subsequently trigger a downstream homeostatic decrease in GSH synthesis.

We note that attempting to explain NAC’s therapeutic effect solely by citing it as a source of cysteine for increased GSH synthesis is likely simplistic and may not hold true in situations where GSH is not severely depleted and the supply of cysteine exceeds the demand, such as in our rats. Indeed, studies have reported NAC-mediated cytoprotection independent of GSH levels (Patriarca et al. 2005) and have proposed that sulfane sulfur species produced as a result of desulfuration of NAC to hydrogen sulfide play important protective roles (Ezerina et al. 2018). Detailed study of NAC’s antioxidative mechanisms is beyond the scope of the current study, and further studies are warranted to test the contribution of an increase (or maintenance) of intracellular GSH concentration to long-term NAC-mediated GSH/GSSG normalization and neuroprotection.

The case for a therapeutic NAC effect thus seems clear regardless of how it is achieved on a molecular level. However, systematic reviews of published clinical literature have not found a consistently favorable NAC effect after TBI and meta-analyses have often been precluded due to variations in sample populations and methodologies, including in dosing regimens and outcome measures (Bhatti et al. 2017). Importantly, one common theme in clinical NAC trials seems to be a relatively short duration of its administration, typically for just a few days after injury (Bhatti et al. 2017). While there may be several reasons why NAC fails to show outcome benefits in humans, our results showing the persistent depression of GSH/GSSG ratios for weeks after injury in rats indicate an inadequate duration of treatment in humans may be one of them.

Bilateral PVI and PNN counts decline progressively following focal TBI and are salvaged by NAC treatment

GSH/GSSG imbalance and resultant cortical oxidative stress culminate in injury to the PNN (Cabungcal et al. 2013a) which in turn may further increase oxidative stress (Hsu et al. 2022), creating a positive feedforward loop that disproportionately affects PVIs over other neurons (Hsieh et al. 2017; Steullet et al. 2017) due to their high baseline metabolic demand (Sohal et al. 2009). Oxidative stress-driven PVI impairment has been reported as a common mechanism in other models of neurocognitive disorders (Steullet et al. 2017), and we previously investigated this aspect by measuring PVI and PNN counts (Hsieh et al. 2017; MacMullin et al. 2020), per-cell PV protein dynamics (MacMullin et al. 2020), and cortical excitability (Hsieh et al. 2017) at timescales of days to weeks after experimental TBI. Our data across species and TBI methods have shown that both PVI and PNN counts progressively decline following injury (Hsieh et al. 2017; MacMullin et al. 2020), and this decline is accompanied by an increase in cortical excitability (Hsieh et al. 2017; Hameed et al. 2019). Here, we show that supplementing endogenous antioxidant reserve immediately after injury preserves both PVI and PNN cell counts in perilesional cortex for up to 6 weeks after injury. This amelioration of PVI and PNN loss may well have therapeutic potential and has been suggested in other neurocognitive diseases (Cabungcal et al. 2013a; Sun et al. 2016) and in our previous TBI work (Hameed et al. 2019).

Our previous work revealed increased extracellular glutamate followed by PNN degradation and PVI loss in a mouse model of repeated mild TBI (MacMullin et al. 2020). This lead to a concern (tested here) that NAC treatment may enhance glutamate efflux into the extracellular space via the Sxc− (Reissner 2014; Soria et al. 2014; Wright et al. 2016) and thus exacerbate PVI injury and PNN degradation - particularly as NAC is a derivative of L-cysteine, which has itself been implicated as an excitotoxin (Janaky et al. 2000). Encouragingly, this is not observed in our results - rather, the PNN is restored upon NAC treatment 3 weeks after injury, while redox reserve is not completely recovered until 6 weeks after TBI. The partial mitigation of oxidative stress coupled with PNN preservation by NAC 3 weeks after TBI may suggest a transient state in which the PNN is recruited for additional antioxidative capacity during recovery.

Perhaps relevant for eventual translation to clinical trials, protection provided by NAC after TBI may be incomplete. TBI triggers multiple pathophysiologic cascades with distinct, time-dependent pathological mechanisms driving secondary injury (Ladak et al. 2019; Ng and Lee 2019; MacMullin et al. 2020), and oxidative stress is thus probably only partially responsible for chronic post-traumatic E:I imbalance. Indeed, our data show contralesional PVI and PNN dropout in the presence of seemingly normal antioxidant capacity in TBIH2O rats both 3 and 6 weeks post-TBI. In addition, while contralesional PVI counts improve, PNN counts do not in TBINAC rats compared to the TBIH2O group 6 weeks after injury despite a normal GSH/GSSG ratio at this timepoint. This may indicate that factors other than oxidative stress contribute to post-TBI changes in cortical biology and E:I ratio, either alone or in conjunction with changes in oxidative stress that are subthreshold for our detection. Further to the potential of subthreshold changes, we note that there is a strong trend toward a TBI-induced decrease in contralesional GSH/GSSG in TBIH2O rats 6 weeks after injury which does not reach predetermined statistical significance. Our findings thus underscore the need for phased, targeted therapeutic strategies as a function of time after injury (Klein et al. 2020).

NAC reverses the transhemispheric decline in per-cell parvalbumin protein expression caused by TBI

Since a vast majority of the cortical PNN surrounds PVIs, as we demonstrated in earlier work (Hsieh et al. 2017), we used this relationship to assess the functional state, indexed by per-cell PV density, of surviving PVIs after injury and asked whether some surviving PVIs are PV-depleted after TBI. Reduced PV content is linked to an impaired ability to sequester intracellular calcium, reduced firing capacity, and impaired inhibition in a paired-pulse synaptic paradigm (Caillard et al. 2000; Wohr et al. 2015). PV-depleted PVIs may thus be limited in terms of their GABAergic inhibitory capacity.

Indeed, we found that injury results not only in a progressive loss of cortical PVIs but also in a delayed reduction in PV expression in those that survive. While the number of PNN-wrapped (PNN+) PVIs expressing undetectable per-cell PV protein levels did not differ between groups 3 weeks after injury, there was a higher percentage of such cells in TBIH2O rats compared to both Sham and TBINAC animals in both hemispheres 6 weeks after TBI, though the cells remain viable as demonstrated by DAPI staining. We hypothesize that these low-PV PNN+ PVIs are likely dysfunctional (Caillard et al. 2000; Schwaller et al. 2004; Wohr et al. 2015), shifting cortical E:I balance further toward excitation.

Interestingly, in addition to preventing PVI loss, NAC treatment also modulates per-cell PV expression within surviving mature PVIs at both timepoints after TBI. Three weeks after injury, NAC treatment induces supranormal PV expression bilaterally within a subset of PNN+ PVIs even as the mode of the distribution shifts in the same direction as in TBIH2O rats. Six weeks after injury, NAC treatment not only prevents the emergence of low-PV PVIs but continues to promote PV overexpression in a subset of these cells. Notably, NAC treatment does not completely restore the GSH/GSSG ratio at the 3 week timepoint, suggesting that PVIs responded to NAC by increasing PV expression before GSH rebalance.

The post-traumatic emergence of low-PV PVIs and their prevention by NAC may hinge on PV expression being activity dependent (Patz et al. 2004). TBI has been demonstrated to reduce feedforward PVI activation in the hippocampus by altering synaptic input and inducing a net inhibition on these cells (Folweiler et al. 2020), and NAC may increase excitatory drive due to both partial agonism of the N-methyl-D-aspartate-type glutamate receptor (NMDAR) by L-cysteine (Janaky et al. 2000) as well as an overall increase in ambient glutamate (Wright et al. 2016). The current results confirm our previous findings (MacMullin et al. 2020), albeit in a different model and species, and further suggest that long-term post-traumatic sequelae might involve PVI “quiescence,” which effectively renders these cells into an inactive state.

NAC-mediated PVI rescue, PNN maintenance, prevention of the low-PV PVI state, and concurrent promotion of PV overexpression in surviving PVIs also provide mechanistic insights into the neuroprotection offered by post-TBI NAC treatment. Enhanced per-interneuron PV expression may compensate for post-traumatic PVI loss. Such fine-tuning of overall cortical PV content suggests that NAC treatment might provide much needed local circuit stability in addition to mitigating oxidative stress, all of which would presumably contribute to limiting secondary injury and chronic post-traumatic sequalae.

Given NAC’s multifaceted role in redox regulation, neuroprotection, and PV modulation, future investigation might focus on whether and how these individual processes are necessary for post-TBI recovery. For example, NMDAR antagonism is neuroprotective but may decrease PV expression (Powell et al. 2012), which might explain why memantine is only partially effective in alleviating post-TBI behavioral symptoms in a mouse model (Mei et al. 2018), and polypharmacy combining NMDAR antagonism and PV overexpression may be a viable and more effective strategy to mitigate post-TBI sequelae and aid recovery.

E‌EG biomarkers of TBI are preserved with NAC treatment

In our experiments aimed at testing NAC treatment's capacity to preserve the PVI system, EEG served as a measure of injury magnitude, as cortical injury reliably increases delta (0.5–4 Hz) EEG power in rodents as well as in humans (Lu et al. 2011; Modarres et al. 2017; Thomasy et al. 2017). TBI caused a significant and sustained increase in delta power in our cohort both 3 and 6 weeks after injury, and NAC administration normalized power to Sham levels at both timepoints.

Given our immunohistochemical results revealing PVI deficits, and the correlation between fast-spiking PVI and gamma EEG bands (Cardin et al. 2009; Sohal et al. 2009; Sohal 2012), we were surprised that no detectable changes occurred in total gamma band (30–90 Hz) power at either timepoint after injury. One explanation for this disconnect may be that locally impacted PVI subtypes, which typically exert local circuit control, are not reflected at the broad EEG activity level. Studies that have reported such correlations used optogenetics (Cardin et al. 2009; Sohal et al. 2009) or diffuse-injury models (MacMullin et al. 2020) which likely engage broader PVI populations. Higher frequencies are also notoriously sensitive to noise during data collection. These EEG frequencies rub up against the spectral bandwidth of muscle activity (Hipp and Siegel 2013; Muthukumaraswamy 2013), and in-vivo studies linking gamma frequencies to PVI mostly measure local field potentials via intracortical electrodes rather than epidural EEG (Cardin et al. 2009; Sohal et al. 2009; Chen et al. 2017). Detecting subtle changes in gamma frequency band signal may be beyond the capabilities of our current recording setup, and follow-up experiments more suited to minimizing electrical and myographic artifacts are planned.

Rather than reduction in gamma EEG power, TBI caused reductions in EEG power in the theta (4–13 Hz) and beta (13–30 Hz) bands, with NAC treatment normalizing these changes to Sham levels. Of note, spontaneous activity in the beta frequency range is associated with somatostatin-positive interneurons (SOMIs), the second most common inhibitory interneuron type (Tremblay et al. 2016; Chen et al. 2017; Veit et al. 2017), which help to coordinate activity across frequency ranges in addition to single-cell level excitability and plasticity control (Booker et al. 2020) and are also lost following TBI.

We note that the notion that GABAergic interneuron subtypes govern specific EEG frequency bands in awake and freely moving animals, while tempting, is simplistic. It is far more probable that these rhythms reflect complex, aggregated interactions between distinct inhibitory and excitatory neurons. For example, although theta activity is studied as a discrete band, it is the integrated result of the outputs of multiple current generators and is influenced by gamma input (Amilhon et al. 2015; Lopez-Madrona et al. 2020). Indeed, PVIs have been reported to be involved in the generation of theta (4–15 Hz) oscillations (Buzsaki 2002; Stark et al. 2013; Amilhon et al. 2015) and the modulation of beta (20–40 Hz) rhythms (Chen et al. 2017). It is also important to note that PVIs form a heterogeneous group with distinct morphological, electrophysiological, and molecular subtypes (Klausberger and Somogyi 2008; Helm et al. 2013; Que et al. 2021), not all of which have been extensively characterized.

The authors also note that sleep disturbance is commonly reported following TBI in humans (Viola-Saltzman and Watson 2012) and both sleep deprivation and hypersomnia can alter spectral EEG content (Cajochen et al. 1999; Hung et al. 2013; Gorgoni et al. 2014; Sasai-Sakuma and Inoue 2015; Fattinger et al. 2017). However, injured rats in our study did not differ from Sham in total time spent in inactive (presumed sleep) states, or in inactive-active transitions. Further, published results indicate that experimental TBI to parietal cortex in rats, while occasionally resulting in mild sleep stage fragmentation, does not lead to changes in total waking or sleep times (Skopin et al. 2015; Buchele et al. 2016; Andrade et al. 2022). We nonetheless propose that NAC-mediated preservation of EEG biomarkers may reflect improved sleep as a result of the rescue of neuronal populations, particularly PVI and SOMI, that regulate sleep and arousal (Skopin et al. 2015; Jones 2016; Scammell et al. 2017; Thankachan et al. 2019; Zucca et al. 2019; McKenna et al. 2020).

Conclusion

In sum, our results highlight the significance of immediate post-TBI restoration of endogenous redox equilibrium via NAC administration as a plausible prophylactic intervention for chronic sequelae related to the loss of cortical inhibitory tone. Further experiments will be necessary to understand the effects of NAC on cortical interneuron physiology and seizure threshold following TBI and whether NAC can be utilized as an adjunct treatment combined with other neuroprotective therapies to improve outcomes. Ultimately, targeting relevant time-dependent pathophysiological pathways after injury via a multipronged approach combining multiple concurrent therapies might be the best therapeutic strategy for post-TBI management.

Supplementary Material

Figure-S1-legend_bhac327
Click here for additional data file. (12.5KB, docx)

Acknowledgements

We thank the IDDRC Animal Behavior and Physiology Core at Boston Children’s Hospital, funded by NIH/NICHD P50 HD105351.

Contributor Information

Mustafa Q Hameed, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Neuromodulation Program, Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Department of Neurosurgery, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States.

Nathaniel Hodgson, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States.

Henry H C Lee, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Rosamund Stone Zander Translational Neuroscience Center, Boston Children’s Hospital, 300 Longwood Avenue, Boston, MA 02115, United States.

Andres Pascual-Leone, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Neuromodulation Program, Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States.

Paul C MacMullin, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Neuromodulation Program, Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States.

Ali Jannati, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Neuromodulation Program, Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States.

Sameer C Dhamne, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Neuromodulation Program, Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States.

Takao K Hensch, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Department of Molecular & Cellular Biology, Center for Brain Science, Harvard University, 52 Oxford Street, Cambridge, MA 02138, United States.

Alexander Rotenberg, F.M. Kirby Neurobiology Center, Department of Neurology, Boston Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Neuromodulation Program, Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, United States; Rosamund Stone Zander Translational Neuroscience Center, Boston Children’s Hospital, 300 Longwood Avenue, Boston, MA 02115, United States.

Funding

This study was supported by the National Institute of Neurological Disorders and Stroke at the National Institutes of Health (grant number: 1R01NS088583 to A.R.); and the Boston Children’s Hospital Translational Research Program to A.R.

Conflict of interest statement: All authors declare that they have no conflicts of interest.

References

  1. Amilhon  B, Huh  CY, Manseau  F, Ducharme  G, Nichol  H, Adamantidis  A, Williams  S. Parvalbumin interneurons of hippocampus tune population activity at theta frequency. Neuron. 2015:86(5):1277–1289. [DOI] [PubMed] [Google Scholar]
  2. Andrade  P, Lara-Valderrabano  L, Manninen  E, Ciszek  R, Tapiala  J, Ndode-Ekane  XE, Pitkanen  A. Seizure susceptibility and sleep disturbance as biomarkers of epileptogenesis after experimental TBI. Biomedicines. 2022:10(5):1138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aoyama  K, Watabe  M, Nakaki  T. Regulation of neuronal glutathione synthesis. J Pharmacol Sci. 2008:108(3):227–238. [DOI] [PubMed] [Google Scholar]
  4. Bhatti  J, Nascimento  B, Akhtar  U, Rhind  SG, Tien  H, Nathens  A, da  Luz  LT. Systematic review of human and animal studies examining the efficacy and safety of N-acetylcysteine (NAC) and N-acetylcysteine amide (NACA) in traumatic brain injury: impact on neurofunctional outcome and biomarkers of oxidative stress and inflammation. Front Neurol. 2017:8:744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bondy  SC, LeBel  CP. The relationship between excitotoxicity and oxidative stress in the central nervous system. Free Radic Biol Med. 1993:14(6):633–642. [DOI] [PubMed] [Google Scholar]
  6. Booker  SA, Harada  H, Elgueta  C, Bank  J, Bartos  M, Kulik  A, Vida  I. Presynaptic GABAB receptors functionally uncouple somatostatin interneurons from the active hippocampal network. Elife. 2020:9:e51156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Buchele  F, Morawska  MM, Schreglmann  SR, Penner  M, Muser  M, Baumann  CR, Noain  D. Novel rat model of weight drop-induced closed diffuse traumatic brain injury compatible with electrophysiological recordings of vigilance states. J Neurotrauma. 2016:33(13):1171–1180. [DOI] [PubMed] [Google Scholar]
  8. Buzsaki  G. Theta oscillations in the hippocampus. Neuron. 2002:33(3):325–340. [DOI] [PubMed] [Google Scholar]
  9. Cabungcal  JH, Steullet  P, Kraftsik  R, Cuenod  M, Do  KQ. Early-life insults impair parvalbumin interneurons via oxidative stress: reversal by N-acetylcysteine. Biol Psychiatry. 2013a:73(6):574–582. [DOI] [PubMed] [Google Scholar]
  10. Cabungcal  JH, Steullet  P, Morishita  H, Kraftsik  R, Cuenod  M, Hensch  TK, Do  KQ. Perineuronal nets protect fast-spiking interneurons against oxidative stress. Proc Natl Acad Sci U S A. 2013b:110(22):9130–9135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Caillard  O, Moreno  H, Schwaller  B, Llano  I, Celio  MR, Marty  A. Role of the calcium-binding protein parvalbumin in short-term synaptic plasticity. Proc Natl Acad Sci U S A. 2000:97(24):13372–13377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cajochen  C, Foy  R, Dijk  DJ. Frontal predominance of a relative increase in sleep delta and theta EEG activity after sleep loss in humans. Sleep Res Online. 1999:2(3):65–69. [PubMed] [Google Scholar]
  13. Cammarota  M, Losi  G, Chiavegato  A, Zonta  M, Carmignoto  G. Fast spiking interneuron control of seizure propagation in a cortical slice model of focal epilepsy. J Physiol. 2013:591(4):807–822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cantu  D, Walker  K, Andresen  L, Taylor-Weiner  A, Hampton  D, Tesco  G, Dulla  CG. Traumatic brain injury increases cortical glutamate network activity by compromising GABAergic control. Cereb Cortex. 2015:25(8):2306–2320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cardin  JA, Carlen  M, Meletis  K, Knoblich  U, Zhang  F, Deisseroth  K, Tsai  LH, Moore  CI. Driving fast-spiking cells induces gamma rhythm and controls sensory responses. Nature. 2009:459(7247):663–667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Chen  G, Zhang  Y, Li  X, Zhao  X, Ye  Q, Lin  Y, Tao  HW, Rasch  MJ, Zhang  X. Distinct inhibitory circuits orchestrate cortical beta and gamma band oscillations. Neuron. 2017:96(6):1403–1418.e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dewan  MC, Rattani  A, Gupta  S, Baticulon  RE, Hung  YC, Punchak  M, Agrawal  A, Adeleye  AO, Shrime  MG, Rubiano  AM, et al.  Estimating the global incidence of traumatic brain injury. J Neurosurg. 2019:130(4):1080–1097. [DOI] [PubMed] [Google Scholar]
  18. Dhamne  SC, Silverman  JL, Super  CE, Lammers  SHT, Hameed  MQ, Modi  ME, Copping  NA, Pride  MC, Smith  DG, Rotenberg  A, et al.  Replicable in vivo physiological and behavioral phenotypes of the Shank3B null mutant mouse model of autism. Mol Autism. 2017:8(1):26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Dixon  SJ, Patel  DN, Welsch  M, Skouta  R, Lee  ED, Hayano  M, Thomas  AG, Gleason  CE, Tatonetti  NP, Slusher  BS, et al.  Pharmacological inhibition of cystine-glutamate exchange induces endoplasmic reticulum stress and ferroptosis. Elife. 2014:3:e02523. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Dorsett  CR, McGuire  JL, Niedzielko  TL, DePasquale  EA, Meller  J, Floyd  CL, McCullumsmith  RE. Traumatic brain injury induces alterations in cortical glutamate uptake without a reduction in glutamate transporter-1 protein expression. J Neurotrauma. 2017:34(1):220–234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Ezerina  D, Takano  Y, Hanaoka  K, Urano  Y, Dick  TP. N-acetyl cysteine functions as a fast-acting antioxidant by triggering intracellular H2S and sulfane sulfur production. Cell Chem Biol. 2018:25(4):447–459.e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Fattinger  S, Kurth  S, Ringli  M, Jenni  OG, Huber  R. Theta waves in children's waking electroencephalogram resemble local aspects of sleep during wakefulness. Sci Rep. 2017:7(1):11187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Faul  M, Coronado  V. Epidemiology of traumatic brain injury. Handb Clin Neurol. 2015:127:3–13. [DOI] [PubMed] [Google Scholar]
  24. Folweiler  KA, Xiong  G, Best  KM, Metheny  HE, Nah  G, Cohen  AS. Traumatic brain injury diminishes feedforward activation of parvalbumin-expressing interneurons in the dentate gyrus. eNeuro. 2020:7(6):ENEURO.0195–ENEU19.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ghiam  MK, Patel  SD, Hoffer  A, Selman  WR, Hoffer  BJ, Hoffer  ME. Drug repurposing in the treatment of traumatic brain injury. Front Neurosci. 2021:15:635483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Godoy  LD, Prizon  T, Rossignoli  MT, Leite  JP, Liberato  JL. Parvalbumin role in epilepsy and psychiatric comorbidities: from mechanism to intervention. Front Integr Neurosci. 2022:16:765324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Goodrich  GS, Kabakov  AY, Hameed  MQ, Dhamne  SC, Rosenberg  PA, Rotenberg  A. Ceftriaxone treatment after traumatic brain injury restores expression of the glutamate transporter, GLT-1, reduces regional gliosis, and reduces post-traumatic seizures in the rat. J Neurotrauma. 2013:30(16):1434–1441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Gorgoni  M, Ferlazzo  F, Ferrara  M, Moroni  F, D'Atri  A, Fanelli  S, Gizzi Torriglia  I, Lauri  G, Marzano  C, Rossini  PM, et al.  Topographic electroencephalogram changes associated with psychomotor vigilance task performance after sleep deprivation. Sleep Med. 2014:15(9):1132–1139. [DOI] [PubMed] [Google Scholar]
  29. Hameed  MQ, Hsieh  TH, Morales-Quezada  L, Lee  HHC, Damar  U, MacMullin  PC, Hensch  TK, Rotenberg  A. Ceftriaxone treatment preserves cortical inhibitory interneuron function via transient salvage of GLT-1 in a rat traumatic brain injury model. Cereb Cortex. 2019:29(11):4506–4518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Helm  J, Akgul  G, Wollmuth  LP. Subgroups of parvalbumin-expressing interneurons in layers 2/3 of the visual cortex. J Neurophysiol. 2013:109(6):1600–1613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hipp  JF, Siegel  M. Dissociating neuronal gamma-band activity from cranial and ocular muscle activity in EEG. Front Hum Neurosci. 2013:7:338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Hsieh  TH, Lee  HHC, Hameed  MQ, Pascual-Leone  A, Hensch  TK, Rotenberg  A. Trajectory of parvalbumin cell impairment and loss of cortical inhibition in traumatic brain injury. Cereb Cortex. 2017:27(12):5509–5524. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Hsu  SY, Chen  CH, Mukda  S, Leu  S. Neuronal Pnn deficiency increases oxidative stress and exacerbates cerebral ischemia/reperfusion injury in mice. Antioxidants (Basel). 2022:11(3):466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Hung  CS, Sarasso  S, Ferrarelli  F, Riedner  B, Ghilardi  MF, Cirelli  C, Tononi  G. Local experience-dependent changes in the wake EEG after prolonged wakefulness. Sleep. 2013:36(1):59–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Janaky  R, Varga  V, Hermann  A, Saransaari  P, Oja  SS. Mechanisms of L-cysteine neurotoxicity. Neurochem Res. 2000:25(9/10):1397–1405. [DOI] [PubMed] [Google Scholar]
  36. Jones  BE. Neuroscience: What are cortical neurons doing during sleep?  Curr Biol. 2016:26(21):R1147–R1150. [DOI] [PubMed] [Google Scholar]
  37. Kelly  E, Schaeffer  SM, Dhamne  SC, Lipton  JO, Lindemann  L, Honer  M, Jaeschke  G, Super  CE, Lammers  SH, Modi  ME, et al.  mGluR5 modulation of behavioral and epileptic phenotypes in a mouse model of tuberous sclerosis complex. Neuropsychopharmacology. 2018:43(6):1457–1465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Khatri  N, Thakur  M, Pareek  V, Kumar  S, Sharma  S, Datusalia  AK. Oxidative stress: major threat in traumatic brain injury. CNS Neurol Disord Drug Targets. 2018:17(9):689–695. [DOI] [PubMed] [Google Scholar]
  39. Klausberger  T, Somogyi  P. Neuronal diversity and temporal dynamics: the unity of hippocampal circuit operations. Science. 2008:321(5885):53–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Klein  P, Friedman  A, Hameed  MQ, Kaminski  RM, Bar-Klein  G, Klitgaard  H, Koepp  M, Jozwiak  S, Prince  DA, Rotenberg  A, et al.  Repurposed molecules for antiepileptogenesis: Missing an opportunity to prevent epilepsy?  Epilepsia. 2020:61(3):359–386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Kochanek  PM, Jackson  TC, Ferguson  NM, Carlson  SW, Simon  DW, Brockman  EC, Ji  J, Bayir  H, Poloyac  SM, Wagner  AK, et al.  Emerging therapies in traumatic brain injury. Semin Neurol. 2015:35(1):83–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ladak  AA, Enam  SA, Ibrahim  MT. A review of the molecular mechanisms of traumatic brain injury. World Neurosurg. 2019:131:126–132. [DOI] [PubMed] [Google Scholar]
  43. Lewerenz  J, Hewett  SJ, Huang  Y, Lambros  M, Gout  PW, Kalivas  PW, Massie  A, Smolders  I, Methner  A, Pergande  M, et al.  The cystine/glutamate antiporter system x(c)(−) in health and disease: from molecular mechanisms to novel therapeutic opportunities. Antioxid Redox Signal. 2013:18(5):522–555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Lopez-Madrona  VJ, Perez-Montoyo  E, Alvarez-Salvado  E, Moratal  D, Herreras  O, Pereda  E, Mirasso  CR, Canals  S. Different theta frameworks coexist in the rat hippocampus and are coordinated during memory-guided and novelty tasks. Elife. 2020:9:e57313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Lu  SC. Regulation of glutathione synthesis. Mol Asp Med. 2009:30(1–2):42–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Lu  SC. Glutathione synthesis. Biochim Biophys Acta. 2013:1830(5):3143–3153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Lu  XC, Hartings  JA, Si  Y, Balbir  A, Cao  Y, Tortella  FC. Electrocortical pathology in a rat model of penetrating ballistic-like brain injury. J Neurotrauma. 2011:28(1):71–83. [DOI] [PubMed] [Google Scholar]
  48. MacMullin  P, Hodgson  N, Damar  U, Lee  HHC, Hameed  MQ, Dhamne  SC, Hyde  D, Conley  GM, Morriss  N, Qiu  J, et al.  Increase in seizure susceptibility after repetitive concussion results from oxidative stress, parvalbumin-positive interneuron dysfunction and biphasic increases in glutamate/GABA ratio. Cereb Cortex. 2020:30(12):6108–6120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. McKenna  JT, Thankachan  S, Uygun  DS, Shukla  C, McNally  JM, Schiffino  FL, Cordeira  J, Katsuki  F, Zant  JC, Gamble  MC, et al.  Basal forebrain parvalbumin neurons mediate arousals from sleep induced by hypercarbia or auditory stimuli. Curr Biol. 2020:30(12):2379–2385.e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Mei  Z, Qiu  J, Alcon  S, Hashim  J, Rotenberg  A, Sun  Y, Meehan  WP  3rd, Mannix  R. Memantine improves outcomes after repetitive traumatic brain injury. Behav Brain Res. 2018:340:195–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Modarres  MH, Kuzma  NN, Kretzmer  T, Pack  AI, Lim  MM. EEG slow waves in traumatic brain injury: convergent findings in mouse and man. Neurobiol Sleep Circadian Rhythms. 2017:2:59–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Moga  D, Hof  PR, Vissavajjhala  P, Moran  TM, Morrison  JH. Parvalbumin-containing interneurons in rat hippocampus have an AMPA receptor profile suggestive of vulnerability to excitotoxicity. J Chem Neuroanat. 2002:23(4):249–253. [DOI] [PubMed] [Google Scholar]
  53. Murphy  TH, Miyamoto  M, Sastre  A, Schnaar  RL, Coyle  JT. Glutamate toxicity in a neuronal cell line involves inhibition of cystine transport leading to oxidative stress. Neuron. 1989:2(6):1547–1558. [DOI] [PubMed] [Google Scholar]
  54. Muthukumaraswamy  SD. High-frequency brain activity and muscle artifacts in MEG/EEG: a review and recommendations. Front Hum Neurosci. 2013:7:138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Ng  SY, Lee  AYW. Traumatic brain injuries: pathophysiology and potential therapeutic targets. Front Cell Neurosci. 2019:13:528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Patriarca  S, Furfaro  AL, Domenicotti  C, Odetti  P, Cottalasso  D, Marinari  UM, Pronzato  MA, Traverso  N. Supplementation with N-acetylcysteine and taurine failed to restore glutathione content in liver of streptozotocin-induced diabetics rats but protected from oxidative stress. Biochim Biophys Acta. 2005:1741(1–2):48–54. [DOI] [PubMed] [Google Scholar]
  57. Patz  S, Grabert  J, Gorba  T, Wirth  MJ, Wahle  P. Parvalbumin expression in visual cortical interneurons depends on neuronal activity and TrkB ligands during an early period of postnatal development. Cereb Cortex. 2004:14(3):342–351. [DOI] [PubMed] [Google Scholar]
  58. Powell  SB, Sejnowski  TJ, Behrens  MM. Behavioral and neurochemical consequences of cortical oxidative stress on parvalbumin-interneuron maturation in rodent models of schizophrenia. Neuropharmacology. 2012:62(3):1322–1331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Purtell  H, Dhamne  SC, Gurnani  S, Bainbridge  E, Modi  ME, Lammers  SHT, Super  CE, Hameed  MQ, Johnson  EL  3rd, Sahin  M, et al.  Electrographic spikes are common in wildtype mice. Epilepsy Behav. 2018:89:94–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Que  L, Lukacsovich  D, Luo  W, Foldy  C. Transcriptional and morphological profiling of parvalbumin interneuron subpopulations in the mouse hippocampus. Nat Commun. 2021:12(1):108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Reissner  KJ. The cystine/glutamate antiporter: when too much of a good thing goes bad. J Clin Invest. 2014:124(8):3279–3281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Ruden  JB, Dugan  LL, Konradi  C. Parvalbumin interneuron vulnerability and brain disorders. Neuropsychopharmacology. 2021:46(2):279–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Rudy  B, Fishell  G, Lee  S, Hjerling-Leffler  J. Three groups of interneurons account for nearly 100% of neocortical GABAergic neurons. Dev Neurobiol. 2011:71(1):45–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Sasai-Sakuma  T, Inoue  Y. Differences in electroencephalographic findings among categories of narcolepsy-spectrum disorders. Sleep Med. 2015:16(8):999–1005. [DOI] [PubMed] [Google Scholar]
  65. Scammell  TE, Arrigoni  E, Lipton  JO. Neural circuitry of wakefulness and sleep. Neuron. 2017:93(4):747–765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Schafer  FQ, Buettner  GR. Redox environment of the cell as viewed through the redox state of the glutathione disulfide/glutathione couple. Free Radic Biol Med. 2001:30(11):1191–1212. [DOI] [PubMed] [Google Scholar]
  67. Schwaller  B, Tetko  IV, Tandon  P, Silveira  DC, Vreugdenhil  M, Henzi  T, Potier  MC, Celio  MR, Villa  AE. Parvalbumin deficiency affects network properties resulting in increased susceptibility to epileptic seizures. Mol Cell Neurosci. 2004:25(4):650–663. [DOI] [PubMed] [Google Scholar]
  68. Sharma  S, Tiarks  G, Haight  J, Bassuk  AG. Neuropathophysiological mechanisms and treatment strategies for post-traumatic epilepsy. Front Mol Neurosci. 2021:14:612073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Skopin  MD, Kabadi  SV, Viechweg  SS, Mong  JA, Faden  AI. Chronic decrease in wakefulness and disruption of sleep-wake behavior after experimental traumatic brain injury. J Neurotrauma. 2015:32(5):289–296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Sohal  VS. Insights into cortical oscillations arising from optogenetic studies. Biol Psychiatry. 2012:71(12):1039–1045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Sohal  VS, Zhang  F, Yizhar  O, Deisseroth  K. Parvalbumin neurons and gamma rhythms enhance cortical circuit performance. Nature. 2009:459(7247):698–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Soria  FN, Perez-Samartin  A, Martin  A, Gona  KB, Llop  J, Szczupak  B, Chara  JC, Matute  C, Domercq  M. Extrasynaptic glutamate release through cystine/glutamate antiporter contributes to ischemic damage. J Clin Invest. 2014:124(8):3645–3655. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Stark  E, Eichler  R, Roux  L, Fujisawa  S, Rotstein  HG, Buzsaki  G. Inhibition-induced theta resonance in cortical circuits. Neuron. 2013:80(5):1263–1276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Steullet  P, Cabungcal  JH, Coyle  J, Didriksen  M, Gill  K, Grace  AA, Hensch  TK, LaMantia  AS, Lindemann  L, Maynard  TM, et al.  Oxidative stress-driven parvalbumin interneuron impairment as a common mechanism in models of schizophrenia. Mol Psychiatry. 2017:22(7):936–943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Sun  XR, Zhang  H, Zhao  HT, Ji  MH, Li  HH, Wu  J, Li  KY, Yang  JJ. Amelioration of oxidative stress-induced phenotype loss of parvalbumin interneurons might contribute to the beneficial effects of environmental enrichment in a rat model of post-traumatic stress disorder. Behav Brain Res. 2016:312:84–92. [DOI] [PubMed] [Google Scholar]
  76. Tenorio  M, Graciliano  NG, Moura  FA, Oliveira  ACM, Goulart  MOF. N-acetylcysteine (NAC): impacts on human health. Antioxidants (Basel). 2021:10(6):967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Thankachan  S, Katsuki  F, McKenna  JT, Yang  C, Shukla  C, Deisseroth  K, Uygun  DS, Strecker  RE, Brown  RE, McNally  JM, et al.  Thalamic reticular nucleus parvalbumin neurons regulate sleep spindles and electrophysiological aspects of schizophrenia in mice. Sci Rep. 2019:9(1):3607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Thomasy  HE, Febinger  HY, Ringgold  KM, Gemma  C, Opp  MR. Hypocretinergic and cholinergic contributions to sleep-wake disturbances in a mouse model of traumatic brain injury. Neurobiol Sleep Circadian Rhythms. 2017:2:71–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Tremblay  R, Lee  S, Rudy  B. GABAergic interneurons in the neocortex: from cellular properties to circuits. Neuron. 2016:91(2):260–292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Veit  J, Hakim  R, Jadi  MP, Sejnowski  TJ, Adesnik  H. Cortical gamma band synchronization through somatostatin interneurons. Nat Neurosci. 2017:20(7):951–959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Viola-Saltzman  M, Watson  NF. Traumatic brain injury and sleep disorders. Neurol Clin. 2012:30(4):1299–1312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Wen  TH, Binder  DK, Ethell  IM, Razak  KA. The perineuronal 'Safety' net? Perineuronal net abnormalities in neurological disorders. Front Mol Neurosci. 2018:11:270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Wilson  L, Stewart  W, Dams-O'Connor  K, Diaz-Arrastia  R, Horton  L, Menon  DK, Polinder  S. The chronic and evolving neurological consequences of traumatic brain injury. Lancet Neurol. 2017:16(10):813–825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Wohr  M, Orduz  D, Gregory  P, Moreno  H, Khan  U, Vorckel  KJ, Wolfer  DP, Welzl  H, Gall  D, Schiffmann  SN, et al.  Lack of parvalbumin in mice leads to behavioral deficits relevant to all human autism core symptoms and related neural morphofunctional abnormalities. Transl Psychiatry. 2015:5(3):e525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  85. Wright  DJ, Gray  LJ, Finkelstein  DI, Crouch  PJ, Pow  D, Pang  TY, Li  S, Smith  ZM, Francis  PS, Renoir  T, et al.  N-acetylcysteine modulates glutamatergic dysfunction and depressive behavior in Huntington's disease. Hum Mol Genet. 2016:25(14):2923–2933. [DOI] [PubMed] [Google Scholar]
  86. Yamamoto  T, Rossi  S, Stiefel  M, Doppenberg  E, Zauner  A, Bullock  R, Marmarou  A. CSF and ECF glutamate concentrations in head injured patients. Acta Neurochir Suppl. 1999:75:17–19. [DOI] [PubMed] [Google Scholar]
  87. Zhuang  Z, Shen  Z, Chen  Y, Dai  Z, Zhang  X, Mao  Y, Zhang  B, Zeng  H, Chen  P, Wu  R. Mapping the changes of glutamate using glutamate chemical exchange saturation transfer (GluCEST) technique in a traumatic brain injury model: a longitudinal pilot study. ACS Chem Neurosci. 2019:10(1):649–657. [DOI] [PubMed] [Google Scholar]
  88. Zitka  O, Skalickova  S, Gumulec  J, Masarik  M, Adam  V, Hubalek  J, Trnkova  L, Kruseova  J, Eckschlager  T, Kizek  R. Redox status expressed as GSH:GSSG ratio as a marker for oxidative stress in paediatric tumour patients. Oncol Lett. 2012:4(6):1247–1253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  89. Zucca  S, Pasquale  V, de  Leon  L, Roig  P, Panzeri  S, Fellin  T. Thalamic drive of cortical parvalbumin-positive interneurons during down states in anesthetized mice. Curr Biol. 2019:29(9):1481–1490.e6. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure-S1-legend_bhac327
Click here for additional data file. (12.5KB, docx)

Articles from Cerebral Cortex (New York, NY) are provided here courtesy of Oxford University Press

RESOURCES