Temporal differential effects of post-injury alcohol consumption in a mouse model of blast-induced traumatic brain injury

Abstract

Traumatic brain injury is a prevalent condition that affects millions worldwide with no clear understanding or effective therapeutic management available. Military soldiers have a high risk of exposure to blast-induced traumatic brain injury (bTBI). Furthermore, alcohol drinking is common in this population, and studies have shown that post-TBI alcohol exposure can result in memory loss. Hence, it is possible that alcohol could contribute to the overall pathological outcome of brain trauma. However, such a possibility has not been explored in detail. Here, we combined a mild bTBI (mbTBI) model with the drinking-in-the-dark (DID) paradigm to investigate the pathological synergy between mbTBI and alcohol consumption by examining brain oxidative stress levels and behavioral alterations in mice. The results revealed the anxiolytic and short-term memory improvement effects of post-trauma alcohol drinking examined at an early timepoint post mbTBI. However, extended alcohol drinking for up to three weeks post mbTBI impaired long-term memory and was accompanied by intensified oxidative stress in brain regions associated with memory and anxiety. These findings, as well as those from previous in vitro TBI/alcohol studies, suggest a pathological synergy of physical force and post-impact alcohol exposure. This knowledge could potentially aid in establishing guidelines for TBI victims to avoid further injury to their brains as well as to help maximize their recovery following TBI.

INTRODUCTION

Traumatic brain injuries (TBI) are highly prevalent in the global population, with an estimated 69 million people suffering from a TBI annually.¹ A specific type of TBI, blast-induced TBI (bTBI), is predominant among military personnel and war-zone citizens.² It is also well established that, similar to other central nervous system traumas, the pathological outcome of bTBI depends on not only the initial physical impact, or primary injury, but also the subsequent harmful chemical cascades, or secondary injury.^3–9 It is evident that the secondary injury could be influenced by the interplay of endogenous pathways, but it could also be exacerbated by exogenous components, deemed “risk factors”.^10–13 It is well documented that even mild TBI could be linked to long-term forms of neurodegeneration likely resulting from ongoing secondary injuries, particularly when the impact is repeated.^{5, 13–15} However, the understanding, identification, and treatment of this debilitating condition has been hampered by the lack of conspicuous initial motor and sensory symptoms in mild bTBI (mbTBI), the most common form of combat bTBI.^{7, 16} This lack of knowledge is thought to delay diagnosis past the prime window for intervention, resulting in active military personnel suffering repeated blast injuries that worsen the trauma with each additional impact until culminating into intractable damage. Therefore, a fundamental goal of this work is to deepen our knowledge of the pathogenic mechanisms, both intrinsic and extrinsic, that serve as contributing factors of bTBI. This would ultimately facilitate early disease detection as well as more effective therapeutic interventions and prevention measures, which will not only benefit patients, but their families and society as a whole.

Most studies on TBI focus on the post-trauma pathogenesis that is directly instigated by the impact. However, less attention has been paid to the risk factors such as alcohol that could serve as key contributing factors to its overall pathological manifestation. The underlying cause of alcohol use after TBI is confounded by many environmental and psychological factors.^17–19 Nevertheless, it is conceivable that their synergistic effects have chronic biological and psychosocial consequences.¹³ The relationship between TBI and substance use has been studied by many, as reviewed by Christopher M. Olsen and John D. Corrigan.²² However, due to the difficulty of conducting a blast-related TBI in a laboratory setting, few studies have investigated the comorbidity of bTBI and alcohol consumption. Thus, it remains uncertain whether alcohol consumption following bTBI can lead to a worsening of short and long-term neuronal damage resulting from mbTBI.

It has been well established that oxidative stress and its associated lipid peroxidation are major pathological contributors in TBI.^{7, 23–26} One of the earliest cellular responses of oxidative stress post TBI is the production of reactive oxygen species (ROS), which leads to lipid peroxidation and the production of aldehydes.^{7, 27, 28} As both a product and catalyst of lipid peroxidation, acrolein and other aldehydes induce a vicious cycle of oxidative stress, dramatically amplifying its effects and continuously propagating both spatial and temporal degeneration in CNS trauma.^{8, 29–42} As the most reactive aldehyde, acrolein is a well-established and widely used marker by which to detect this oxidative stress.^{32, 34, 43–51} In particular, acrolein has been shown to be a key contributor to oxidative stress and serves as a critical pathological biomarker in TBI.^{8, 36, 52, 53} Thereby, changes in acrolein levels are an established indication of oxidative stress following mbTBI.

While others have examined alcohol drinking after repeated mbTBI,⁵⁴ few have used a free alcohol-drinking protocol that is incorporated with behavioral testing. Therefore, we decided to employ the drinking-in-the-dark (DID) protocol in a mouse mbTBI model integrated with behavioral testing to assess motor, memory, and anxiety functions.^{55, 56} We believed that such a combination of a more clinically-relevant alcohol drinking model in addition to multiple behavioral tests, along with acrolein detection in multiple brain regions correlated with those behaviors, could unveil critical information on the mechanisms of bTBI in general, and the role of alcohol in TBI specifically.

Using this unique combination of an established bTBI and alcohol drinking model, along with well-known behavioral and biochemical analysis, we have demonstrated that consuming alcohol following mbTBI in mice could restore locomotive activity and short-term memory in addition to mitigating anxiety-like behavior in the subacute stage (1 week post mbTBI and DID). However, the continuous consumption of alcohol beyond the subacute stage (3 weeks post mbTBI and DID) resulted in long-term memory deficits and oxidative stress in brain regions known to be linked to memory and anxiety-like behavior. Such findings could provide critical knowledge that not only serves to deepen our understanding of the mechanisms of TBI, but also helps to establish effective diagnostic, treatment, and preventive strategies.

EXPERIMENTAL PROCEDURES

Animals and mild blast model

Twenty-four male C57Bl/6J (C57BL/6NHsd) mice (Envigo Inc.) aged 2.5~3 months (weight 26– 31 g; mean 28.8 ± 1.7 g) were used in this study (12 mice were given mbTBI, and 12 mice were given sham exposure) (Fig. 1E). Only male mice were used to reflect the demographic profile in the military population and to reduce animal waste. All animal procedures were conducted under animal use protocols approved by the Purdue University Animal Care and Use Committee (PACUC). Mice were single-housed in a room with a reverse light-dark cycle (light-off: 10am~10pm; light-on: 10pm~10am). The blast injury, alcohol consumption, and all behavioral experiments were conducted in the dark during the light-off timeframe. The bTBI shock tube model was adapted from a rat model as described in prior publications (Fig. 1B,C).^{7, 8, 57, 58} The blast tube consists of a top driver section (height: 7cm) and a bottom driven section (height: 54cm) with an inner diameter of 5cm, separated by a mylar membrane. When the nitrogen gas starts filling in the driver section, the mylar membrane collapses and creating a gas wave traveling down the driven section and impacting the animal’s head.

Figure 1.

Experimental design and timeline.

(A) An overview of the drinking-in-the-dark (DID) experimental timeline. Mice were acclimated to the drinking routine with water for 2 weeks before the start of the experiment. They were subjected to either blast (bTBI groups) or sham injury (Sham groups) on day 0. Neurological Severity Scores (NSS) were evaluated on day 1 prior to the DID period, which began on day 2. Each week during the DID period, either 20% alcohol or water was provided to the animals for 2 hours for 3 days and for 4 hours on the 4th day to simulate binge drinking. Two sets of behavioral tests were performed at 1 week and 3 weeks post injury, including an Open Field Test (OFT), two Novel Object Recognition Tests (NOR), and a Light Dark Box Test (LDB).

(B) A schematic of the shock tube blast device. Mouse was placed in prone position beneath the blast tube. A mylar membrane was inserted in the interface between the top driver and bottom driven section, and nitrogen was fill from the driver chamber until the membrane erupt, creating an explosion wave traveling down the blast tube hitting the dorsal part of the animal’s head.

(C) A representative blast wave recording. The blast wave was recorded using a pressure transducer and an oscilloscope.

(D) A schematic of the DID setup. All mice were on reverse light/dark cycles. At 1pm, the home cage water bottles were switched with drinking bottles that contained either a 20% alcohol solution or water for the mice to drink. After each drinking session, alcohol drinking bottles were removed and replaced with the original water bottles. The volume of alcohol consumed was documented from the bottles.

(E) The four treatment groups and their respective sample sizes used in this study. 24 mice were randomly divided into 4 groups based on their treatments: 1. Sham-H2O: Sham (blast sound only) with water; 2. bTBI-H2O: Blast injury with water; 3. Sham-EtOH: Sham with 20% alcohol; 4. bTBI-EtOH: Blast injury with 20% alcohol. Each group began with 6 animals. However, one mouse died in the bTBI-EtOH group after the 1st week of drinking.

Abbreviations: bTBI—blast traumatic brain injury; NSS—Neurological Severity Score; OFT—Open Field Test; NOR—Novel Object Recognition Test; LDB—Light Dark Box Test; EtOH—20% alcohol solution.

In brief, a pair of mice were anesthetized together with isoflurane in a gas chamber (induction: 4%; maintenance: 2~3%). The sham mouse was removed from the gas chamber and placed in an open container on a desk next to the blast device. The mbTBI mouse was transferred to a platform in a prone position with its head resting on a polyurethane foam under the blast nozzle. The rest of its body was protected by a custom-made acrylic shield. The blast wave transmitted from the dorsal to ventral side of the mouse’s head. The pressure at the surface of the mouse was measured to be around 75 kPas (~11 PSI). Three consecutive blasts were done in one day with an approximately 3-minute inter-blast interval.

Righting reflex and modified-Neurological Severity Score (m-NSS)

Immediately following exposure to the final blast wave, both sham and mbTBI mice were placed in the supine position. The time it took for the animals to regain their righting reflex and flip back to the prone position was measured. This serves as an indication for acute loss-of-consciousness experienced by each mouse following injury. Afterwards, mice were returned to their home cages for recovery.

A revised neurological severity score for mice was used as described previously, ⁵⁹ with modifications. Twenty-four hours post blast/sham injury, mice were habituated in the behavioral room for at least 30 minutes. A transparent cage (Cage A) and a 0.5 cm-wide balance beam were set up. Mice were removed from their home cages and separately tested in Cage A. The m-NSS includes 10 tasks rated on a scale of 20 to measure the mouse’s sensorimotor responses. The tests included: 1. General Balance; 2. Landing Reflex; 3. Tail Raise Reflex; 4. Drag Response; 5. Righting Reflex; 6. Ear Reflex; 7. Eye Reflex; 8. Sound Reflex; 9. Foot Reflex; and 10. Tail Reflex. Following the examination of each animal, the testing cage and balance beam were cleaned with 70% ethanol. The mice were returned to their home cages immediately after testing.

Procedure for drinking in the dark (DID)

A four-day drinking in the dark (DID) protocol was used based on the procedure reported previously. ⁵⁵ All mice were trained to use custom drinking bottles with water for two weeks prior to the start of the DID experiment. During the first three days of DID, the home cage water bottles were removed 3 hours after the start of the dark cycle and were either replaced with water or 20% alcohol in the custom drinking bottles for 2 hours (1pm~3pm) (Fig. 1A,D). Subsequently, the custom drinking bottles were removed and the home cage drinking bottles were returned to each cage respectively. The amounts consumed from the drinking bottles were determined by observing the volume change before and after the drinking session. On the fourth day, mice were allowed to binge drink for 4 hours (1pm~5pm).

Open field test (OFT)

The open field test (OFT) was conducted at 1 week and 3 weeks post injury. Only the first OFT conducted at the 1 week timepoint was used for anxiety-like behavior analysis due to concerns about complex effects associated with repeated assessments ^60–62, whereas both OFTs at 1 week and 3 weeks were used for locomotor assessment. Mice were placed in the middle of a 40 cm × 40 cm square open field with black walls and a white floor. The field was in a dark room illuminated by a dim red overhead light. The duration of the test was 10 minutes per animal. The percentage of time animals spend in the center vs. peripheral zones in the open field is used as an indication of anxiety-like behavior towards a novel environment, where less time spent in the center corresponds to a potentially more anxious state.⁶³ Following the test, mice were returned to their home cages. The open field apparatus was cleaned with 70% ethanol after the testing of each animal. Mice movements were tracked and analyzed using ANY-Maze software.

Novel object recognition test (NOR)

The novel object recognition test was a two-day test conducted the day after OFT at 1 week and 3 weeks post injury (Fig. 1A). During the familiarization phase on day one, two identical objects (Objects A) were placed in the same 40 cm × 40 cm open field for each mouse under the same illumination settings. Each mouse was allowed to explore the objects for 5 minutes. Two hours after familiarization, one of the identical objects was switched to a novel one (Object B) in the same location, and mice were again allowed to explore the objects for 5 minutes to test for short-term memory. On day two (i.e. 24 hours after the familiarization phase), Object B was then replaced with a different novel object (Object C) in the same location, and mice were allowed to explore the objects for 5 minutes to test for long-term memory. After each of the above tests, mice were placed back in their home cages. The open field apparatus was cleaned with 70% ethanol between the testing of each animal. Mice movements were tracked and analyzed using ANY-Maze software. Discrimination index (DI) was calculated as the time spent exploring the novel object divided by the total time exploring both objects. The Objects A, B, and C used in the 1 wpi timepoint were completely different from those used at the 3 wpi time point.

Light-dark box test

The light-dark box test was conducted at 3 weeks post injury (Fig. 5C,D). The light-dark box consisted of a light compartment and a smaller dark compartment. Bright white light was used to illuminate the light compartment. Mice were placed in the middle of the light compartment and allowed to explore for 5 minutes. After each test, mice were returned to their home cages and the light-dark box was cleaned with 70% ethanol. The result of this test was expressed as the percentage of time spent in the dark compartment divided by the total exploration period. Mouse movements were tracked and analyzed using ANY-Maze software.

Figure 5.

Assessment of anxiety-like behavior at 1 week and 3 weeks post mbTBI/alcohol drinking.

(A,B) Open field test to assess anxiety-like behavior at 1 week post injury. The values of % of time in the center zone for the groups of Sham-H2O, bTBI-H2O, Sham-EtOH, bTBI-EtOH are 9.56 ± 0.41, 7.44 ± 0.76, 11.10 ± 1.25, and 14.66 ± 1.95 respectively. The data shows that mbTBI alone caused a trend of less time spent in the center of the field, although not significant, when compared to Sham (bTBI-H2O vs Sham-H2O, p > 0.05). However, mbTBI mice that consumed alcohol did increase the amount of time spent in the center of the field compared to mbTBI without alcohol intake (bTBI-EtOH vs bTBI-H2O, ** p = 0.0022, t = 4.313, df = 19). Furthermore, mbTBI mice that consumed alcohol even spent more time in the center zone than Sham mice without alcohol intake (bTBI-EtOH vs Sham-H2O, * p = 0.0396, t = 3.049, df = 19).

(C,D) Light-dark box test at 3 weeks post injury. At 3 weeks post injury, no differences in the amount of time spent in the light vs. dark zone were detected among the four experimental groups.

All data were analyzed with two-way ANOVA and Bonferroni test. All values are expressed as mean ± SEM. N = 6 for each group except for bTBI-EtOH group with n = 5 due to one mouse dying in cage at the end of week 1.

Immunohistochemistry (IHC) staining

At the end of the study, mice were anesthetized with a ketamine/xylazine (100/10 mg/kg) mixture and transcardially perfused with ice-cold phosphate buffered saline (PBS) followed by 4% paraformaldehyde (PFA) fixation. Mouse brains were extracted and post-fixed in 4% PFA at 4°C overnight. Subsequently, brains were dehydrated with 15% sucrose and then 30% sucrose for at least 72 hours. The brains were then immersed in OCT and frozen in dry ice. Coronal sections of 25 μm thickness were sliced using a cryostat. Ten sections were cut around bregma −2.0mm according to anatomical landmarks shown in the Allen Brain Institute Mouse Brain Atlas. The sections were stored in a cryoprotectant solution at −20°C as described in a prior publication.⁶⁴ For each animal, at least 3 tissue sections 25 μm apart were used for immunohistology staining. Selected sections were removed from the cryoprotectant solution and washed in PBS 3 × 5 minutes. This was followed by a 30-minute incubation in 3% H₂O₂ in water, a 20-minute perforation in 1.5% Triton in PBS, and a 1.5 hour blocking in 10% normal goat serum in PBS at room temperature. Tissues were then incubated overnight in mouse anti-acrolein antibody (1:1000, StressMarq) with gentle agitation at 4 °C. The next day, tissues were rinsed in 0.1% Triton in PBS for 3 × 10 minutes and incubated in goat anti-mouse biotinylated secondary antibody (1:1000, Vector Laboratory) for 2 hours at room temperature. Following another wash in 0.1% Triton in PBS for 3 × 10 minutes, tissues were incubated in ABC Peroxidase Standard Staining Kit (Thermofisher) for 30 minutes. Tissues were then transferred to 0.1% Triton in PBS for 3 × 5 minutes before incubating in Pierce^™ DAB Substrate (Thermofisher) for 6 minutes. The reaction was quenched in di-water and tissues were mounted onto slides to air-dry overnight. Slides were further dehydrated with 2 minutes of 50%−70%−95%−100%−100% ethanol, and then 2 × 3 minutes of Xylene. Finally, slides were coverslipped with Fisher Chemical^™ Permount^™ Mounting Medium for imaging analysis.

Image acquisition and analysis

All DAB-stained sections were imaged using an Olympus IX51 microscope and a CMOS Color Camera with a 4x and 10x objective. Individual images were saved in a RGB TIFF file format using the CellSens software (Olympus Corporation). Image processing and analysis were conducted using the ImageJ software. Original images were processed using [Color Deconvolution—H DAB], specific regions of interests (ROIs) were selected using the Polygon Selection Tool, and mean gray values were measured for each ROI. Regions in both hemispheres were analyzed. The selection of brain regions was based on the acrolein adduct accumulation results from the lab’s previous pilot experiments. The IHC procedure and analysis method for detecting acrolein adducts has been reported in prior studies.^{35, 65–67}

Statistical analysis

Based on prior studies, a sample size of 6 mice per group was chosen to achieve statistical significance for the parameters of interest ⁵². All statistical analyses were performed using GraphPad Prism 4.0 (GraphPad Software, Inc.). All data are presented as Mean ± Standard Error of Mean (SEM). Unpaired t-test was used in the measurement of righting reflex, m-NSS score, and alcohol consumption. For all other data, two-way ANOVA was used to compare differences among the four groups and p values for multiple comparisons were adjusted using Bonferroni’s test. Outlier analysis was performed using the Grubbs’ Test with significance level α=0.05. Homoscedasticity plots were used to confirm equal variance. Differences were considered significant at p < 0.05 for all analyses, with * indicating p < 0.05; **, p < 0.01; ***, p < 0.001; and ****, p < 0.0001.

RESULTS

Acute functional loss following blast injury

To determine the acute functional deficits caused by blast injury, we examined the righting reflex and modified-neurological severity scores of each mouse. The time for each sham or mbTBI mouse to regain their righting reflex served as an indication of acute loss-of-consciousness. This was measured immediately following the last blast injury. As expected, mbTBI mice required a significantly longer time to regain their righting reflex when placed in the supine position compared to the sham injury group (***p = 0.0006) (Fig. 2A). Twenty-four hours after the blast injury, a modified neurological severity score (m-NSS) was obtained to determine the injury severity.⁵⁹ Overall, blasted animal displayed higher m-NSS scores than their sham counterparts (p < 0.0001) (Fig. 2B). This is indicative of a significant sensorimotor impairment at the acute timepoint post injury.

Figure 2.

Neurological deficits of injured mice at 1 day post mbTBI.

(A) Comparison between the Sham and bTBI mice righting reflexes, expressed as the amount of time mice took to regain their prone position after been placed in the supine position immediately after blast/sham injury. The values are as follows: 25.75 ± 4.99 s (Sham) and 84.50 ± 13.74 s (bTBI). Note that mbTBI resulted in a significant increase in time to regain an upright position when compared to Sham (*** p = 0.0006, t = 4.019, df = 22, unpaired t-test). N = 12 for each group.

(B) Mouse neurological severity scores (m-NSS) measured at 1 day post Sham/bTBI. Blasted mice showed significantly higher m-NSS scores (2.33 ± 0.26) compared to the Sham group (0.083 ± 0.08) (**** p <0.0001, t = 8.350, df = 22, unpaired t-test). N = 12 for each group. Tests of both A and B were conducted prior to any alcohol consumption, so the Sham-H2O and Sham-EtOH, as well as bTBI-H2O and bTBI-EtOH groups were pooled together for this analysis. All values, A and B, are expressed as mean ± SEM.

mbTBI and Sham mice have no significant differences in alcohol drinking throughout three weeks of DID

To determine whether blast injury affects the amount of alcohol consumed by mice, we measured the volume (ml) of alcohol drunk by the mice over the course of the three drinking-in-the-dark (DID) periods. As indicated in Figure 3, there was a slight tendency for the bTBI-EtOH mice to drink more than the Sham-EtOH group during the binge drinking periods in week 1 and 3. However, the differences in drinking volumes (ml) and alcohol intake (g/kg body weight) were not statistically significant between any groups (p > 0.05).

Figure 3.

Alcohol consumption over 3 weeks of drinking-in-the-dark (DID).

(A) Total volume of alcohol consumed in Sham (Sham-EtOH group) and mbTBI (bTBI-EtOH group) mice over three weeks of DID. (p > 0.05 for both binge drinking day and average 2-hr drinking days at week 1, 2, or 3; two-way ANOVA)

(B) Alcohol intake expressed as the drinking volume normalized by body weight (ml/kg) for the Sham (Sham-EtOH group) and mbTBI (bTBI-EtOH group) mice over three weeks of DID. (p > 0.05 for both binge drinking day and average 2-hr drinking days at week 1, 2, or 3; two-way ANOVA)

Data in all cases represent the mean ± SEM. N = 6 for each group except for bTBI-EtOH group at week 2 & 3 with n = 5 due to one mouse dying in cage before week 2. There were no statistically significant differences between the groups of Sham-EtOH and bTBI-EtOH at each time point. ).

Subacute motor deficits in blasted mice at one week post injury can be alleviated by alcohol consumption

To evaluate locomotor activities following blast injury, an open field test (OFT) was used at 1 week and 3 weeks post injury. At 1 week post injury/DID, mbTBI mice given only water (bTBI-H₂O) traveled less distances in the open field compared to their sham counterparts (Sham-H₂O) (* p = 0.0201) (Fig. 4A). Similarly, the mice in the bTBI-H₂O group traveled with slower speeds compared to those in the Sham-H₂O group (* p = 0.0207) (Fig. 4B). However, no difference in movements was observed between bTBI-H₂O and bTBI-EtOH in terms of distance (p > 0.05) (Fig. 4A) or speed (p > 0.05) (Fig. 4B). In addition, no difference was observed between Sham-H₂O and bTBI-EtOH in terms of distance (p > 0.05) or speed (p > 0.05) (Fig. 4B). This indicates that alcohol did not significantly impact motor deficits following mbTBI despite a tendency for it to mitigate motor impairment following injury. Furthermore, at 1 week post mbTBI, as a result of alcohol consumption, the significance of motor deficits in bTBI compared to Sham-H2O disappeared. At 3 weeks post injury/DID, no motor deficits were observed in any group of mice (Fig. 4C,D, p > 0.05) in the open field.

Figure 4.

Motor function assessments in various experimental groups 1 week and 3 weeks post mbTBI/alcohol drinking.

(A,B) Open-field assessment at 1 week post injury. The values of travel distance (A) for the groups of Sham-H2O, bTBI-H2O, Sham-EtOH, bTBI-EtOH are 27.76 ± 1.40, 20.72 ± 1.93, 25.33 ± 2.18, and 25.29 ± 1.28 respectively. The values of travel speed (B) for the groups of Sham-H2O, bTBI-H2O, Sham-EtOH, bTBI-EtOH are 0.046 ± 0.002, 0.034 ± 0.003, 0.042 ± 0.004, and 0.042 ± 0.002 respectively. Note that the bTBI-H2O mice traveled significantly less distance (*p = 0.0201, t = 2.86, df = 19) (A) and with slower speeds (*p = 0.0207, t = 2.849, df = 19) (B) in the open field compared to Sham-H2O mice. However, alcohol consumption eliminated these deficits in travel distance (in A, bTBI-EtOH vs Sham-H2O: p > 0.05), as well as travel speeds (in B, bTBI-EtOH vs Sham-H2O: p > 0.05) compared to Sham-H2O following mbTBI.

(C,D) Open-field assessment at 3 weeks post injury. At this timepoint, there were no significant differences among the four experimental groups in terms of the distance traveled (all comparisons p > 0.05).

All data were analyzed with two-way ANOVA with Bonferroni test. All values are expressed as mean ± SEM. N = 6 for each group except for bTBI-EtOH group with n = 5 due to one mouse dying in cage at the end of week 1.

Alcohol consumption promotes exploratory behavior in blasted mice at one week post-injury

To test if animals developed anxiety-like behaviors following mbTBI, the percentage of time mice spent in the center zone in the open field at 1 week post injury was measured. Animals that exhibit anxiety-like phenotypes towards a novel environment are known to spend less time in the center of an open zone.⁶³ Although the bTBI-H₂O mice showed a tendency of spending less time in the center zone compared to the Sham-H₂O mice, there was no statistical difference between these two groups ( p > 0.05) (Fig. 5 A,B). However, there was a significant difference between the bTBI-EtOH and bTBI-H₂O group (** p = 0.0022). No significant difference was noted between bTBI-EtOH and Sham-EtOH (p > 0.05). Interestingly, bTBI-EtOH mice also spent more time in the center zone compared to the Sham-H₂O mice (* p = 0.0396 ) (Figure 5A).

To assess anxiety-like behavior at a later stage post-injury, a light-dark box test was performed 3 weeks after the injury/DID (Fig. 5C,D). As indicated in Figure 5C, no differences were observed in the four experimental groups, suggesting that mild blast injury and/or alcohol consumption did not lead to significant changes in anxiety-like behavior at this time point (p > 0.05, Fig. 5C,D).

Short-term and long-term memory changes at one week post TBI and alcohol consumption

To investigate whether the blast injury and/or alcohol drinking affected mouse object recognition memory function at 1 week post injury, a novel object recognition test was conducted at 1 week post injury (Fig. 6 A–D). All mice were first given a similar amount of time to explore two identical objects during the familiarization stage (Fig. 6A). Then, one object was switched to a novel one two hours later (as a test of short-term memory). All Sham-H₂O mice spent more time exploring a novel object than a familiar one (Figure 6C, DI > 0 for all Sham-H₂O). This was in contrast to the bTBI-H₂O mice that spent approximately equal amounts of time exploring both the novel and familiar objects (DI ≈ 0 for bTBI- H₂O), resulted in a significant decrease in DI (* p = 0.0363) (Fig. 6C).. Surprisingly, blasted mice that consumed alcohol (bTBI-EtOH) had significantly higher discrimination indices than those that only consumed water (bTBI-H₂O) (* p = 0.0241) (Fig. 6C). No statistical differences were observed in long-term memory at one week post injury among various groups (p > 0.05) (Fig. 6D).

Figure 6.

Short-term and long-term object memory assessments at 1 week and 3 weeks post mbTBI/alcohol drinking.

Memory status is expressed as the discrimination index (DI), which is calculated as the percentage of time spent with a novel object divided by the total time of object exploration. B, E represents quantification during the familiarization period, while C, F represents the analyses for the short-term memory test and D, G the long-term memory test.

(A) A schematic of the novel object recognition (NOR) test. The test consists of a familiarization step in which mice were presented with two identical objects, followed by a short-term memory test 2 hours after familiarization in which one of the objects was switched to a new one. On the following day, long-term memory was tested by switching the previously novel object for a new one. This test was conducted at both 1 week and 3 weeks post injury, with a different set of objects used for each week.

(B-D) Short- and long-term memory assessments at 1-week post injury. At 1 week post injury, no differences in DI were detected among the various groups when two identical objects were presented during the familiarization period (B, p > 0.05). In C, one object was switched to assess the short-term memory retrieval task. bTBI- H₂O group had a significant decrease of DI compared to Sham- H₂O group (* p = 0.0363, t = 2.585, df = 19). Additionally, blasted animals who consumed alcohol spent more time with the novel object compared to injured animals who only consumed water (bTBI-EtOH vs. bTBI-H2O, * p = 0.0241, t = 2.775, df = 19). There were no significant differences in long-term object memory retrieval among the various experimental groups (D, p > 0.05).

(E-G) Short- and long-term memory assessments at 3-week post injury. Similarly, at 3 weeks post injury, no differences in DI were detected among the various groups when two identical objects were presented during familiarization (E, p > 0.05). Additionally, no differences were detected among the various groups during short-term memory retrieval tasks (F, p > 0.05). In G, as indicated, no differences were detected when the Sham-H2O group was compared to bTBI-H2O groups (p > 0.05), or to bTBI-EtOH (p > 0.05). However, blasted animals with three weeks of alcohol consumption had significantly lower levels of discrimination indices for the long-term memory test compared to the Sham group with alcohol intake (bTBI-EtOH vs Sham-EtOH, ** p = 0.0059, t = 3.405, df = 19).

All data were analyzed with two-way ANOVA with Bonferroni test. All values are expressed as mean ± SEM. N = 6 for each group except for bTBI-EtOH group with n = 5 due to one mouse dying in cage at the end of week 1.

Short-term and long-term memory changes at three weeks post TBI and alcohol consumption

To determine if blast injury and chronic alcohol drinking had effects on short and long-term memory beyond one week post injury, we repeated the NOR test at 3 weeks post injury. As expected, mice showed no preferences when exploring identical objects during the familiarization period (p > 0.05) (Fig. 6E). Two hours later, no differences were observed in the mice’s short-term memory retrieval abilities (p > 0.05) (Fig. 6F). However, blasted mice that consumed alcohol (bTBI-EtOH) showed a statistically significant deficit in their long-term memory performance compared to their sham counterparts when tested 24 hours after the initial familiarization period (Sham-EtOH) (** p = 0.0059) (Fig. 6G). No difference was observed between the Sham-H₂O and bTBI-H₂O groups.

Oxidative stress marker acrolein was elevated in multiple brain regions at three weeks post injury and alcohol drinking

At three weeks post injury, we found that acrolein adducts were elevated in the bTBI-EtOH mice compared to the Sham-H₂O group in the retrosplenial cortex (** p = 0.0044) (Fig. 7A–C), medial amygdala (* p = 0.0232) (Fig. 7D–F), and CA1 region of the hippocampus (* p = 0.0436) (Fig. 7G–I). We also measured acrolein adduct intensity in other brain regions such as the sensorimotor cortex or basolateral amygdala, but did not find any significant elevation in these regions (data not shown). In all cases of the three aforementioned regions, although either bTBI-H₂O or Sham-EtOH alone showed a tendency of elevated acrolein compared to Sham-H₂O, no significance was reached for these statistical comparisons (p > 0.05). No significant increase in acrolein staining was found in the dentate gyrus for any groups (p > 0.05).

Figure 7.

Changes of acrolein-lysine adducts in brain regions at 3 weeks post mbTBI/alcohol drinking.

Brain tissues were harvested at the end of the study and acrolein-modified proteins were detected using DAB staining. Regions related to memory and anxiety were chosen for examination. Dashed rectangles indicate the location of the zoomed inset, and solid line rectangles show the magnified inset highlighting acrolein adduct accumulation.

(A-I) Acrolein-lysine modified proteins are significantly elevated in the retrosplenial cortex, medial amygdala, and CA1(stratum oriens) regions 3 weeks after blast injury + alcohol consumption.

(A,B,C) Representative immunofluorescence staining of acrolein-lysine adducts in the retrosplenial cortex for the Sham-H2O (A) and bTBI-EtOH groups (B). Bar graph depicts the quantification of the acrolein-lysine adduct staining (C). As indicated in C, acrolein-lysine adducts are significantly elevated in bTBI-EtOH when compared to Sham-H2O (30.82 ± 1.05 vs. 23.96 ± 1.31, ** p = 0.0044, t = 4.023, df = 19).

(D,E,F) Representative immunofluorescence staining of acrolein-lysine adducts in the medial amygdala for the Sham-H2O (D) and bTBI-EtOH groups (E). Bar graph depicts the quantification of the acrolein-lysine adduct staining (F). As indicated in F, acrolein-lysine adducts are significantly elevated in bTBI-EtOH when compared to Sham-H2O (31.52 ± 1.57 vs. 24.66 ± 2.12, * p = 0.0232, t = 3.288, df = 19).

(G,H,I) Representative immunofluorescence staining of acrolein-lysine adducts in the CA1 region for Sham-H2O (G) and bTBI-EtOH group (H). Bar graph depicts the quantification of the acrolein-lysine adduct staining (I). As indicated in I, acrolein-lysine adducts are significantly elevated in bTBI-EtOH when compared to Sham-H2O (29.11 ± 1.13 vs. 24.05 ± 1.49, * p = 0.0436, t = 4.072, df = 19).

(J,K,L) Representative immunofluorescence staining of acrolein-lysine adducts in the dentate gyrus for Sham-H2O (J) and bTBI-EtOH group (K). Despite a slight elevation of staining demonstrated in bTBI-EtOH compared to Sham-H2O, no significance was detected through quantification (p > 0.05) (L). All data were analyzed with two-way ANOVA with Bonferroni test. All values are expressed as mean ± SEM. N = 5 or 6 per group in all cases. Scale bar in the large image is 200μm and scale bar in the inset is 100μm.

DISCUSSION

In this investigation, we observed that, although there was a tendency of elevated alcohol drinking, mice suffering from a mbTBI did not demonstrate a significant difference in alcohol intake compared to their sham counterparts throughout the experimental period. The DID protocol has previously been applied to a mouse repetitive mild controlled cortical impact and a rat mild blast TBI model, which also showed no difference in the dose or preference of alcohol drinking between groups, despite their decrease in cognitive behavioral performance and increased cytokine expression.^{68, 69} This suggests that, similar to the current model of mbTBI and alcohol exposure, mbTBI alone does not lead to higher levels of alcohol consumption. In addition to the similar amounts of alcohol intake between the injured and sham groups, we also noted that alcohol consumption in sham mice did not cause any significant changes in behavioral functions or brain biochemistry when tested. However, alcohol drinking did alter the behavior and biochemistry of mice that were subjected to a mild bTBI. Specifically, at one week post injury and alcohol consumption, mbTBI resulted in a significant impairment of motor functions. However, such a difference was not present when alcohol was consumed. On the other hand, mbTBI alone did not lead to significant changes in anxiety-like behavior, but integrating alcohol with mbTBI (bTBI-EtOH) caused injured mice to display less anxiety-like behavior and more exploratory behavior than the mice in the bTBI-H2O or Sham-H2O groups 1 week post mbTBI and alcohol drinking. These phenomena were time-dependent, however, as at three weeks post injury and alcohol drinking, the injury-mediated motor deficits and alcohol-induced over-exploration in injured mice subsided. This open field test result is in stark contrast with the results reported in some focal mild TBI-alcohol comorbidity studies, such as in Sophie X Teng et al.,⁷⁰ where a significant decrease in % time in the center was observed in the TBI/Alcohol group after mild lateral fluid percussion injury. The difference in mice’s open field behavior after TBI/Alcohol exposure highlights a potential distinction in underlying pathologies between focal physical impact TBI versus diffuse blast TBI, despite their similar injury severity levels. Such a difference could be due to a variation of brain regions involved in different TBI/alcohol consumption models, and further studies need to be conducted to elucidate the exact mechanism. Concerning memory assessment, no short-term and long-term deficits were observed due to mbTBI at both one week and three weeks post injury. Interestingly, bTBI-EtOH mice displayed stronger short-term memory capabilities than bTBI-H2O mice when examined one-week post injury and alcohol drinking. In contrast, bTBI-EtOH mice showed lesser long-term memory capabilities than Sham-EtOH when assessed at three weeks post injury and alcohol drinking. In addition to behavioral testing, biochemical examination showed significantly elevated levels of acrolein adducts, an indicator of oxidative stress, in the bTBI-EtOH group, but not the bTBI-H2O or Sham-EtOH groups, compared to the Sham-H2O group at three weeks post injury and alcohol exposure. Taken together, these findings indicate a pathologically synergistic role of prolonged alcohol consumption when combined with mbTBI in promoting brain oxidative stress and some behavior abnormalities related to anxiety-like behavior and memory.

While we have observed behavioral changes in addition to acrolein elevations in animals administered alcohol following TBI, the exact molecular mechanisms linking alcohol to altered behavior remains to be explored. It is well established that alcohol has a complex influence on anxiety-like behavior and memory changes, partially due to its inhibition of glutamate receptors as well as its ability to downregulate glutamate transporters.^{72, 73, 76–78} These receptors play important roles in regulating synaptic plasticity, and the dysregulations of the receptors may lead to complex behavioral outcomes.^{71, 74} In addition, it is possible that toxic aldehydes such as acrolein, resulting from lipid peroxidation, as well as acetaldehyde, could arise from alcohol exposure to synergistically damage neuronal tissues and contribute to the behavioral changes observed in this study.⁷⁵

In the current study, while blasted mice spent a significantly longer time to regain their righting reflex compared to the sham group (84.5 s vs. 25.8 s, Figure 2A), a value of 85 s in the current mbTBI model is considered mild and is considerably shorter than that in a severe blast model, which is approximately 500 s.⁷⁹ Similarly, among the injured mice in this study, the average value of m-NSS, another measure of injury severity in TBI studies with comprehensive sensorimotor criterion,⁵⁹ is 2.3. This value is significantly higher than that in sham mice at 24 hours post injury (~0.08), and corresponds to a ‘mild’ injury according to the literature.⁸⁰ In conclusion, multiple established TBI measures have confirmed the mild nature of our repeated blast injury model.

In many alcohol studies using rodents, researchers deploy administration methods such as forced injections or oral gavage to standardize the alcohol intake amount. However, these methods can induce unwanted stress in mice resulting from the manner in which the alcohol is administrated,^{81, 82} which could complicate the interpretation of the effects of alcohol. In contrast to forced alcohol intake, the drinking-in-the-dark (DID) protocol is an effective, voluntary, and non-invasive method to study alcohol drinking in mice in their home cage, with minimal stress associated with the administration of alcohol.⁵⁵ In the current study, we have shown that alcohol consumption did affect the stress and anxiety-like behavior of mice following mbTBI. Based on the aforementioned reasoning, we are confident in concluding that any observed changes in stress resulting from alcohol consumption can be attributed to the alcohol itself and was not influenced by the way in which it was administered. Using the DID model, we also observed that there was no significant difference in the volume of alcohol intake between the sham and mbTBI mice throughout the experimental timeframe (Fig. 3A,B). This result suggests that despite not manually controlling the alcohol intake, the amount of total alcohol consumption for each animal remained relatively similar. Therefore, it is reasonable to conclude that the behavioral and histological outcomes are not confounded by any variation in the amounts of alcohol consumed, nor the way in which the alcohol was administered, but are solely a result of the alcohol and its alleged synergistic effect with mild bTBI.

From measuring the animals’ movement in an open field post injury, we observed that while mbTBI alone initially reduces a mouse’s activity and exploration in an open field one week post injury, the addition of alcohol led to the disappearance of this reduction. Specifically, at 1 week post injury (1wpi), bTBI-H₂O mice had significantly reduced total distances traveled and average speeds compared to the Sham-H₂O mice (Fig. 4A,B). However, the injured mice with alcohol consumption (bTBI-EtOH group) traveled total distances and speeds that were not different from those in the Sham-H₂O group (Fig. 4A,B). It is worth noting that alcohol alone did not affect the locomotor activities of sham mice significantly (Sham-EtOH vs. Sham-H₂O) (Fig. 4A,B). Taken together, we speculate that the reduction of locomotion at 1wpi is likely due to increased anxiety-like behavior resulting from the injury, which was not present after alcohol intake.⁸³

Consistent with the observation that alcohol nullified the injury-induced locomotion deficits at 1wpi, we observed that mice given alcohol following mbTBI (bTBI-EtOH) displayed significantly less anxiety-like behavior than the mbTBI mice that consumed water (bTBI-H₂O) in the open field test during the same time period (Fig. 5A). Specifically, at 1wpi, the bTBI-EtOH mice spent significantly more time in the center zone compared to both the bTBI-H₂O and even Sham-H₂O mice (Figure 5A). This result implies that the combination of bTBI with alcohol drinking led the mice to be more exploratory than not only the injured mice, but also the sham mice. In general, our study suggests that a more reckless behavior is produced when alcohol is combined with TBI in this case. Since the sham mice that consumed similar amounts of alcohol did not show any significant differences in anxiety-like behavior compared to the Sham-H₂O mice (Fig. 5A), it appears that alcohol only induces this form of reckless behavior in injured mice, but not in healthy mice.

In addition to anxiety, memory impairment is also an important sequela of bTBI in both mice^{84, 85} and humans.^{86, 87} To assess the mice’s short-term and long-term memory, we conducted repeated novel object recognition (NOR) testing at 1 and 3 weeks post injury. At 1 wpi, bTBI-H₂O mice tended to have slightly lower discrimination indices than Sham-H₂O mice, demonstrating a trend of reduced short-term memory (Fig. 6C, Sham-H₂O vs. bTBI-H₂O p = 0.07). Surprisingly, while alcohol consumption in sham mice (Sham-EtOH) produced a similar tendency of reduced short-term memory function, mbTBI mice who drank alcohol (bTBI-EtOH) had significantly higher discrimination indices than their mbTBI counterparts that only drank water. This implies that while alcohol may have a tendency of impairing the short-term memory of healthy mice, it improves the short-term memory function of mbTBI mice at the 1wpi (Fig. 6C). Intriguingly, a recent clinical study discovered that mild alcohol consumption reduced the risk of dementia, whereas heavy drinking increased that risk.⁸⁸ This suggests that whether alcohol could offer beneficial or detrimental effects may depend on the quantity and duration of its consumption.

To further investigate its neuropsychological influence with a drinking period beyond 1 week, we examined alcohol drinking for three weeks post mbTBI. We observed that the short-term memory differences produced by alcohol at 1 week post injury were no longer significant at the chronic timepoint (Fig. 6F). Furthermore, we discovered that bTBI-EtOH mice had significantly lower discrimination indices (DI) than the Sham-EtOH group at 3 weeks post mbTBI/DID (Fig. 6G). Taken together, these results indicate that while one week of alcohol consumption in mice following a mbTBI could have some beneficial effects on short-term memory retrieval, its continued consumption could potentially lead to detrimental effects on long-term memory.

In addition to behavioral impairments, we also discovered intensified acrolein staining at the brain regions of the retrosplenial cortex, hippocampus, and amygdala, areas known to be associated with memory function and anxiety-related behaviors. Specifically, we detected a significant elevation of acrolein adducts in these regions 3 weeks following the combination of blast injury and alcohol drinking (Fig. 7A–I). As a highly reactive aldehyde known as both a product and instigator of oxidative stress and inflammation^89–91, acrolein serves as a multifaceted marker to help correlate biochemical dysfunction with the behavioral deficits observed in this study^{29, 30, 44}. It is known that the retrosplenial cortex has been implicated in object recognition memory⁹² and is connected to the hippocampus to facilitate memory function.⁹³ Additionally, the stratum oriens region of the hippocampal CA1 layer is important for stress-induced deficits in both spatial memory⁹⁴ as well as in long-term object memory.⁹⁵ Finally, the medial amygdala is a region closely related to social anxiety⁹⁶ and responses to aversive stimuli.⁹⁷ Taken together, the evidence of acrolein elevation in brain regions known to be linked to memory and anxiety is in good agreement with, and therefore further supports, the noted behavioral changes. Due to such a correlation of biochemical and behavioral abnormalities, it is reasonable to speculate that acrolein could potentially be a biomarker to indicate and even predict brain functional deficits.

One interesting phenomenon that particularly suggests a synergistic effect of mbTBI and alcohol consumption is the biochemical assessment of acrolein in various mouse conditions. Specifically, while acrolein production tended to be higher, though not significantly, in mouse brain regions linked to memory and anxiety following either a mbTBI or alcohol drinking, a combination of these two factors did result in a significant elevation of acrolein compared to controls. This data is consistent with the findings recently obtained using an in vitro model of concussive TBI, TBI-on-a-chip, which investigated alcohol exposure following physical impact.⁴⁵ Similar to the findings derived from the current in vivo model, the results of the in vitro model also indicate that neither a mild impact nor a brief exposure of ethanol resulted in significant increases of acrolein production in murine neuronal cultures. However, a combination of these treatments led to a significant surge in acrolein. Therefore, the consistency of the synergistic effects between physical force and post-impact alcohol exposure in both in vitro and in vivo TBI/alcohol models has further validated and reinforced the notion that alcohol intake could be a risk factor that potentially worsens the pathology triggered by TBI. However, considering the short-term beneficial effects detected from post-TBI alcohol consumption in our studies (Fig. 4,5,6), as well as other studies reporting neuroprotection from alcohol drinking following TBI,^{98, 99} it is clear that the exposure duration, frequency, and concentration of the alcohol may be the crucial factors in determining these different outcomes.

Conclusion

In summary, the results from this study demonstrate the effects of alcohol drinking on motor, anxiety-like behavior, and short-term memory at one and three weeks post mbTBI. Specifically, extended alcohol drinking up to three weeks impaired the long-term memory of these mice, which was accompanied by oxidative stress in brain regions associated with memory and anxiety. Furthermore, the data from the current in vivo model, as well as those from previous in vitro TBI/alcohol models, all imply a pathological synergy of physical force and post-impact alcohol exposure. This suggests that alcohol intake, particularly if consumed for an extended period, could be a risk factor that potentially exacerbates the pathology triggered by TBI. This knowledge, upon further validation through clinical studies, could potentially help to establish guidelines for TBI victims to avoid further injuring their brains and maximize their recovery following TBI.