Brain, Behavior, and Immunity 60 (2017) 304–311

Contents lists available at ScienceDirect

Brain, Behavior, and Immunity

journal homepage: www.elsevier .com/locate /ybrbi

Full-length Article

Binge ethanol in adulthood exacerbates negative outcomes followingjuvenile traumatic brain injury

http://dx.doi.org/10.1016/j.bbi.2016.11.0090889-1591/� 2016 Elsevier Inc. All rights reserved.

Abbreviations: TBI, traumatic brain injury; DCX, doublecortin; Iba-1, ionizedcalcium binding adaptor molecule 1; TLR4, toll like receptor 4; IL-1b, interleukin 1beta.⇑ Corresponding author at: Biomedical Research Tower #618, 460 West 12th

Ave., Columbus, OH 43210, USA.E-mail address: Ekaterina.karelina@osumc.edu (K. Karelina).

Kate Karelina ⇑, Kristopher R. Gaier, Maya Prabhu, Vanessa Wenger, Timothy E.D. Corrigan, Zachary M. WeilDepartment of Neuroscience, Group in Behavioral Neuroendocrinology and Center for Brain and Spinal Cord Repair, Ohio State University Wexner Medical Center, Columbus,OH 43210, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 5 August 2016Received in revised form 4 October 2016Accepted 8 November 2016Available online 12 November 2016

Keywords:Traumatic brain injuryAlcoholAxon degenerationMicrogliaNeuroinflammationLearning and memoryNeurogenesis

Traumatic brain injuries (TBI) are a major public health problem with enormous costs in terms of healthcare dollars, lost productivity, and reduced quality of life. Alcohol is bidirectionally linked to TBI as manyTBI patients are intoxicated at the time of their injury and we recently reported that, in accordance withhuman epidemiological data, animals injured during juvenile development self-administered signifi-cantly more alcohol as adults than did sham injured mice. There are also clinical data that drinking afterTBI significantly reduces the efficacy of rehabilitation and leads to poorer long-term outcomes. In order todetermine whether juvenile traumatic brain injury also increased the vulnerability of the brain to thetoxic effects of high dose alcohol, mice were injured at 21 days of age and then seven weeks later treateddaily with binge-like levels of alcohol 5 g/kg (by oral gavage) for ten days. Binge-like alcohol produced agreater degree of neuronal damage and neuroinflammation in mice that sustained a TBI. Further, micethat sustained a juvenile TBI exhibited mild learning and memory impairments in adulthood followingbinge alcohol and express a significant increase in hippocampal ectopic localization of newborn neurons.Taken together, these data provide strong evidence that a mild brain injury occurring early in life rendersthe brain highly vulnerable to the consequences of binge-like alcohol consumption.

� 2016 Elsevier Inc. All rights reserved.

1. Introduction

Traumatic brain injuries (TBI) are a major public health problemwith millions of injuries and tens of thousands of deaths occurringannually in the US alone (Langlois et al., 2006). Further, the long-term health consequences and economic costs in terms of healthspending and loss of productivity are staggering (Coronado et al.,2012).

Alcohol use is tightly linked to traumatic brain injuries as alco-hol intoxication is linked to, by some estimates, more than half ofall traumatic brain injuries (Tagliaferri et al., 2006). Much of theresearch into this relationship therefore has focused on the roleof alcohol as a risk factor for TBI, and intoxication at injury as avariable in functional outcomes and recovery (Berry et al., 2011;Chen et al., 2012; Opreanu et al., 2010; Pandit et al., 2014). How-ever, there is also some experimental and clinical evidence that

TBI itself could serve as an independent risk factor for substanceabuse issues, particularly when injuries occur early in development(Bjork and Grant, 2009; Corrigan et al., 2013; Weil et al., 2016a).Most epidemiological studies have reported that although thereis a population of patients that abstain from alcohol after TBI thereis also a substantial number that resume drinking in the monthsafter injury (Bombardier et al., 2003; Kreutzer and Harris, 1990;Ponsford et al., 2007). There is some limited, though controversial,evidence that intoxication at the time of TBI can be neuroprotec-tive (Berry et al., 2011; Chen et al., 2012; Opreanu et al., 2010;Pandit et al., 2014), but alcohol abuse following TBI is clearly prob-lematic as it produces significantly poorer outcomes including:impairing the efficacy of rehabilitation programs; reducing voca-tional opportunities; and substantially increasing the risk for psy-chiatric disorders, post-traumatic seizures, and additional injuries(Corrigan, 1995; Corrigan et al., 2014; Jorge et al., 2005; Vaaramoet al., 2014a,b).

Recently, we reported that female, but not male, mice thatreceived a single mild traumatic brain injury at 21 days of ageself-administered significantly more alcohol in adulthood thandid sham-injured mice or those injured in adulthood (Weil et al.,2016b). These data revealed that injuries occurring at a critical

K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311 305

developmental time point in pre-adolescent mice resulted in long-lasting consequences that contributed to voluntary alcohol abusein adulthood.

Although it is clear that drinking after traumatic brain injuryreduces the effectiveness of rehabilitation therapy and producespoorer outcomes, the precise mechanisms that link drinking toimpaired neuronal recovery and function in this clinical populationare not fully understood. However, there is mounting evidencefrom both the experimental and clinical literature that heavy orbinge-like drinking, even in the absence of other insults, inducesneuroinflammation and degeneration (Alfonso-Loeches et al.,2010; Crews et al., 2011; Hayes et al., 2013; Qin and Crews,2012; Vetreno and Crews, 2012). Further, TBI itself induces longlasting impairments in neuronal structure and function and alsoinduces neuroinflammatory responses. Additionally, inflammatoryevents early in life appear to prime the neuroimmune system toexhibit greater inflammatory responses to other stimuli, such ashigh-dose alcohol, later in life (Bilbo et al., 2010). Thus, heavydrinking by TBI survivors may impair functional outcomes, at leastin part, by inducing inflammatory responses and further impairingalready dysfunctional neuronal circuits.

Our previous study using the self-administration paradigmallowed us to investigate injury-induced alterations in alcoholreward processing, but it did not allow us to investigate the func-tional consequences of alcohol consumption after injury becausesham-injured mice did not voluntarily self-administer high dosesof ethanol (Weil et al., 2016a). Thus, in order to determine whetherheavy drinking after TBI exacerbates the behavioral, neuropatho-logical, and inflammatory consequences of juvenile injuries infemale mice, we administered ten daily binge-like doses of ethanolvia oral gavage to adult females that had sustained a TBI or shaminjury in early life. The goals of this study were to 1) determinewhether juvenile TBI increased the functional deficits and tissuedamage induced by binge-like levels of alcohol and 2) to begin touncover the mechanisms that link TBI, binge-like alcohol, and tis-sue damage. We hypothesized that binge-like exposure to alcoholwould exacerbate tissue damage and functional deficits in previ-ously injured female mice in part by inducing neuroinflammationand impairing neuroplasticity.

2. Methods

2.1. Animals

All procedures were conducted on female Swiss-Webster micederived from breeders purchased from Charles River (Wilmington,MA) and bred at the Ohio State University. Pups were weaned at21 days of age into a standard mouse cage (32 � 16 � 12 cm) withad libitum access to food and filtered tap water. All animals werehoused in a 14:10 light-dark cycle. All procedures were approvedby the OSU Institutional Animal Care and Use Committee, and wereconducted in accordance with the National Institute of Healthguidelines.

2.2. Traumatic brain injury

On the day of weaning, 21 day-old female mice sustained a mildclosed-head injury or a sham injury as described previously (Weilet al., 2016b). All mice were anesthetized with inhaled isoflurane,secured in a stereotaxic frame, and a round 2 mm impactor (ImpactOne device, Leica Biosystems, Richmond IL) was placed on the sur-face of the exposed skull (�1 mm AP, �1 mm ML). TBI mice under-went an impact at 3 mm/s (dwell time 30 ms) to a depth of 1 mm,while sham mice were exposed to an equivalent amount of anes-thesia in the absence of impact. The skin was closed with nylonsuture and mice were returned to their home cages.

2.3. Alcohol administration

Beginning seven weeks following TBI or sham injury, all micewere administered 5 g/kg of a 31.5% solution of ethanol or watercontrol (dosing based on (Bertola et al., 2013)). Ethanol or waterwas administered via daily oral gavage for 10 consecutive days.Mice were weighed daily to account for body mass fluctuationsin the dosing. This paradigm resulted in 4 groups (sham/water,TBI/water, sham/ethanol, TBI/ethanol), which were divided into 3separate cohorts. One cohort was used for assessment of patho-physiology only (n = 8–10 per group), one for microglial extraction(n = 6–8 per group), and another cohort underwent behavioraltesting followed by histological assessment (n = 9–12 per group).See Fig. 1 for experimental timeline.

2.4. Blood ethanol concentration

Blood ethanol concentration (BEC) was assessed via a colori-metric 96-well plate assay (modified from Prencipe et al., 1987).Blood was drawn from the retro-orbital sinus 90 min after the finaloral gavage. Blood samples were centrifuged and serum stored at�80 �C until use. A BEC assay buffer (100 mM KH2PO4; 100 mMK2HPO4; 0.7 mM 4-aminoantopyrine; 1.7 mM chromotropic aciddisodium salt; 50 mg/L EDTA; 50 mL/L triton X-100) was used todilute ethanol standards and serum and all wells were assessedin duplicate. A reaction mixture of alcohol oxidase (Sigma A2404)and peroxidase from horseradish (Sigma P8375) diluted in BECassay buffer was added to all wells and the plate was read at600 nm.

2.5. Histology

Axon damage was assessed via silver staining using theNeuroSilver kit (FD Neurotechnologies, PK301) per manufacturer’sinstructions. Mice were overdosed with sodium pentobarbital(200 mg/kg) and transcardially perfused with 4% paraformalde-hyde. Following an overnight postfix and cryopreservation, brainsections throughout the forebrain were sliced on a cryostat,(40 lm) and collected into a 24-well culture plate for free-floating immunohistochemistry and silver staining. For immuno-histochemistry, tissue was washed with 0.1 M phosphate-buffered saline, quenched in hydrogen peroxide, and incubated inprimary antibody overnight (rabbit anti-Iba1, Wako 019-19741;goat anti-doublecortin, Santa Cruz sc-8006). Biotinylated sec-ondary antibodies (goat anti-rabbit and donkey anti-goat, VectorLaboratories BA-1000 and BA-5000) were applied the followingday, and staining was visualized using the avidin biotin-diaminobenzidene staining method (Vector Laboratories PK-6100and SK-4100).

2.6. Histological analysis

Axonal degeneration (silver staining) was assessed qualitativelyas previously reported (Weil et al., 2014). Briefly, representativetissue throughout the entire forebrain were assessed and scoredon a 4-point scale (0 = few axonal profiles, 3 = dense axonal degen-eration throughout white matter tracts bilaterally).

Microglial cells (Iba1-positive staining) were assessed in theprefrontal cortex. Morphological analysis was conducted usingNeurolucida software (MicroBrightfield). Briefly, 2–3 cells perhemisphere from each brain were selected using strict criteria aspreviously reported (Karelina et al., 2016) and traced for subse-quent sholl analysis. Using NeuroExplorer (MicroBrightfield) con-centric sholl rings spaced 5 lm apart were centered on the cellbody, and the number of intersections, total process length, andcell body surface area were obtained for each cell. Total microglial

Fig. 1. Experimental timeline. All animals underwent TBI or sham injury at 21 days of age and were returned to their home cages for seven weeks. In adulthood, all mice wereadministered ethanol or water via oral gavage daily for ten consecutive days. These studies were conducted on three cohorts, cohort 1 was perfused for tissue collectionimmediately after the final gavage, cohort 2 was used for microglia isolation, and cohort 3 underwent behavioral testing before tissue collection. Passive avoidance testingwas begun twenty-four hours following the final gavage and brain tissue collection was conducted the day after testing was completed.

306 K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311

numbers were counted in a 0.26 mm2 region of interest using Ima-geJ software (Abràmoff et al., 2004) and converted to cell number/mm2.

Doublecortin-positive neurons were counted in the hippocam-pus dentate gyrus, sections spanning approximately �1.94 mmthrough �2.18 mm from bregma. Total cell counts were collectedin the granule cell layer in both the top and bottom blades of thedentate, as well as the molecular layer and dentate hilus.

2.7. Microglial isolation

Microglia were extracted from bilateral frontal lobes of shamand injured mice 24 h following their final oral gavage. Freshbrains were collected, olfactory bulbs were removed and tissuewas dissected (approximately bregma 0.56 mm through3.56 mm) and homogenized in Dulbecco’s phosphate buffer solu-tion (DPBS). The homogenate was centrifuged and the pellet resus-pended in 70% isotonic Percoll (GE Healthcare Life Sciences #17-0891-01). A discontinuous Percoll gradient was created by layering50% and 35% Percoll and topping off with (DPBS). The gradient wasthen centrifuged and the microglia-containing layer in the 50%/70%interface was collected and washed.

2.8. qPCR

Extracted microglia were lysed and cDNA synthesized using theSuperScript III CellsDirect cDNA synthesis kit (Invitrogen#11739010 and #18080200) according to the manufacturer’s pro-tocol. A TaqMan 18S ribosomal RNA primer/probe set (AppliedBiosystems #4319413E) was used as an internal control. Primer/probe sets for the target genes were also purchased at AppliedBiosystems (TLR4: Mm00445273_m1; IL1b: Mm00434228_m1)and amplification was performed on an ABI 7500 Sequencing FastSystem using the TaqMan Fast Advanced Master Mix (#4444963).The universal two-step RT-PCR cycling conditions used were: 50 �Cfor 2 min, 95 �C for 20 s, followed by 40 cycles of 95 �C for 3 s and60 �C for 30 s. Gene expression was analyzed relative to a standardcurve and normalized to the sham/water group.

2.9. Passive avoidance

Learning and memory was evaluated via the passive avoidancetask. The task consisted of a 2-chambered testing environment(one chamber was illuminated and the other was dark) with a gridfloor and guillotine door separating the chambers. On training day,mice were placed individually into the illuminated chamber for1 min; the door was then lifted and remained open until the mouse

stepped into the dark chamber. Once the mouse entered the darkchamber, the door closed, a brief shock (1 mA, 2 s) was deliveredto the grid floor, and the mouse was immediately replaced intoits home cage. Twenty-four hours following the training, eachmouse was again individually placed into the illuminated chamberfor 1 min. Once the guillotine door was open, latency to enter thedark chamber was recorded (with an upper limit of 300 s). Noshock was administered on testing day.

2.10. Statistical analysis

Qualitative analysis of silver staining was assessed via the non-parametric Mann-Whitney U and Kruskall-Wallis tests. Significantoverall Kruskall-Wallis results were followed up by pairwise com-parisons. Microglial sholl analysis was conducted using a repeatedmeasures ANOVA. All histology analyses (cell counts, processlength, cell body area), behavioral data, and gene expression datawere assessed via a 2-way ANOVA (injury � ethanol administra-tion). All statistics were analyzed using the IBM SPSS V.23 soft-ware. All significant overall ANOVA results were followed by theLeast Significant Differences posthoc test. Data were consideredsignificant for p-values <0.05.

3. Results

3.1. Binge-like alcohol administration exacerbates TBI-associatedpathology

Blood ethanol concentrations were assessed on the final day oforal gavage (90 min following the gavage) and indicate a high levelof ethanol in the blood (�0.4 g/dL) which did not differ betweensham and TBI mice (Fig. 2A). In order to determine the pathologicalimpact of high-dose alcohol exposure, we examined whetherbinge-like alcohol levels further exacerbate axonal degenerationin previously injured mice (Fig. 2B–C). As expected, TBI sustainedin pre-adolescence led to a significant degree of axonal degenera-tion (U = 14.0, p < 0.05). However, a 10-day period of binge-likealcohol administration in adulthood significantly exacerbated axondamage in mice that had sustained a TBI as juveniles (H3 = 20.009,p < 0.05). The overall significant Kruskall-Wallis test was followedup by pairwise comparisons which indicated significant differencesbetween the sham/water and TBI/water groups, as well a signifi-cant increase in degeneration in the TBI/ethanol mice comparedto TBI/water (all p < 0.05).

We further examined microglial morphology as a measure ofneuroinflammation. Brain tissue from all four groups wasprocessed for Iba1 immunohistochemistry. Sholl analysis of

Fig. 2. Binge alcohol administration exacerbates TBI-associated pathology. A) Bloodethanol concentrations in sham and TBI mice 90 min after a gavage. B) Analysis ofsilver staining in the forebrain white matter reveals that juvenile TBI-induced axondegeneration is further exacerbated by repeated high dose administration ofalcohol in adulthood. C) Representative staining in the genu of the corpus callosum(bregma 0.50 mm). Scale bar = 50 lm, an asterisk (⁄) indicates statistically signif-icant differences between indicated groups (p < 0.05).

K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311 307

representative microglia in the prefrontal cortex revealed asignificant effect of injury history on microglial morphology(F1,141 = 194.942, p < 0.01) such that mice that had sustained aTBI exhibited microglia with fewer sholl ring intersections(reduced process complexity), indicative of an activated state(Fig. 3A). TBI also significantly reduced overall microglial processlength compared to sham injury (F3,145 = 3.346, p < 0.05; Fig. 3C).An analysis of the total number of microglia in the prefrontal cor-tex revealed main effects of both injury (F1,29 = 6.294, p < 0.05) andethanol treatment (F1,29 = 80,959, p < 0.01), as well as an interac-tion (F1,29 = 8.382, p < 0.01) such that the ethanol-administeredTBI mice exhibited the greatest number of cortical microglia(Fig. 3D). Finally, a measure of microglial cell body area revealeda main effect of injury (F1,27 = 8.879, p < 0.001), and an interaction(F1,27 = 5.427, p < 0.05) such that ethanol significantly increasedcell body area in TBI mice compared to all other groups (Fig. 3E).

3.2. Microglia from injured mice treated with alcohol show increasedpro-inflammatory gene expression

Given that both TBI and alcohol resulted in activated microglia,and the increased total number of microglia in the TBI/ethanolgroup, we examined the microglia directly in order to determinewhether the activational state corresponds to increased neuroin-flammation. Both Toll-like receptor 4 (F3,16 = 3.516, p < 0.05; Fig. 4A)and interleukin-1 beta (F3,19 = 3.311, p < 0.05) gene expression weresignificantly increased in the TBI/ethanol group relative to all othergroups assessed.

3.3. High doses of alcohol impair learning and memory in previouslyinjured mice

Learning and memory was assessed via the passive avoidancetest in all four groups beginning 24 h after the final ethanol (orwater) administration day. This test takes advantage of the fear-motivated tendency of mice to avoid brightly illuminated areas.

Latency to enter the dark chamber on the training day wasrecorded and indicates no significant difference between groups(Fig. 5A). However, analysis of the latency to enter the dark cham-ber (in which each mouse had previously endured a painful shock)revealed an interaction of injury history and ethanol treatment(F1,34 = 5.247, p < 0.05). These data indicate that the TBI/ethanolgroup exhibited a significantly shorter latency to enter theshock-associated chamber compared to the TBI/water group.Importantly, this difference is not likely to be a result of increasedanxiety in the TBI/ethanol group as all mice had similar latencies toescape the light chamber on the training day.

3.4. Binge alcohol and TBI interact to induce ectopic expression ofdoublecortin-positive cells in the hippocampus

Given the learning/memory impairment observed in the TBI/ethanol mice, we conducted an immunohistochemical analysis ofdoublecortin in the hippocampus. Doublecortin (DCX) is tran-siently expressed in progenitor cells committed to the neuronallineage and serves as an indirect marker of hippocampal neuroge-nesis. DCX is expressed for approximately 2–3 weeks in newbornneurons before it is replaced by mature neuronal markers(Brown et al., 2003), thus brain tissue fixed 13 days after the startof the ethanol administration paradigm is likely to contain a sub-stantial subset of neurons developing during our experimentalmanipulation. Here, we assessed DCX-positive cells in the dentategyrus (Fig. 6). The total number of DCX-positive cells in the dentategyrus granule cell layer did not differ significantly by injury orethanol manipulations. However, we identified a significantincrease in the ectopic expression of DCX-positive neurons inethanol-administered TBI mice. Specifically, the TBI/ethanol grouphad significantly more DCX-positive neurons in the dentate hilus(F1,33 = 7.129, p < 0.05) than any other group. Moreover, amongmice that sustained a TBI, binge ethanol significantly increasedDCX expression in the molecular layer (T14 = 2.420, p < 0.03).

4. Discussion

Binge-like levels of alcohol produce significant neuropathology,neuroinflammation, and functional deficits in mice that had expe-rienced a traumatic brain injury during juvenile development.Specifically, alcohol administration led to greater axonal degener-ation and microglial reactivity in mice that had been previouslyinjured. Further, microglia isolated from previously injured ani-mals exhibited significantly increased gene expression for bothTLR4 and IL-1b following alcohol administration. The combinationof binge-like levels of alcohol and TBI also impaired normal neuro-genesis, resulting in increased ectopic expression of DCX-positivecells in the hippocampus. Finally, the combination of high-dosealcohol and juvenile TBI produced impairments in learning andmemory as assessed in the passive avoidance paradigm. Takentogether, these data indicate that a history of traumatic braininjury significantly alters the neural and functional responses tobinge-like levels of alcohol and that the deleterious consequencesof high levels of alcohol in human populations may be related toinflammatory responses and impairments in neural plasticity.

Imaging and post-mortem studies have revealed pronouncedneuroinflammation and degeneration, particularly of white matter,in the brains of alcoholics. Specifically, there is evidence of ongoinginflammation, microglial activation, and cortical and subcorticalatrophy (Harper et al., 2003; He and Crews, 2008; Ikegami et al.,2003). The neurotoxicity of alcohol varies across the lifespan, butis most prominent in cases of long term heavy drinking. Similarly,axonal degeneration is a key pathological marker of TBI, howeverthis pathology resolves over time following a mild injury. Here

Fig. 3. Binge alcohol and TBI alter microglial morphology. A) Sholl analysis in the prefrontal cortex reveals that both alcohol and TBI reduce microglial process complexity,indicative of an activated state, as seen in B) representative images, scale bar = 50 lm. C) TBI also significantly reduced process length. D) The total number of microglia weresignificantly increased in ethanol-treated mice that sustained a juvenile TBI. E) The cell body area of cortical microglia were also significantly increased in ethanol-treated TBImice. An asterisk (⁄) indicates statistically significant differences between the indicated groups (p < 0.05).

Fig. 4. TBI and alcohol administration increase proinflammatory gene expression. Relative mRNA gene expression is shown relative to the sham/water group and indicatesthat A) TLR4 and B) IL-1b is exacerbated by alcohol in mice that underwent juvenile TBI. An asterisk (⁄) indicates statistically significant differences between the indicatedgroups (p < 0.05).

308 K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311

we report that chronic exposure to binge-like levels of alcohol con-tribute to sustained axonal damage even when the drinking occursmany weeks after the actual injury. Importantly, the binge-likealcohol administration paradigm used in this study was not suffi-cient to induce axonal degeneration in mice that had not experi-enced a prior TBI; rather TBI renders axons vulnerable to thetoxic effects of high dose alcohol.

Mild traumatic brain injury, such as that used in the currentstudy, is characterized by a significant but transient inflammatoryresponse and a number of cognitive and emotional disruptions thatoccur in the acute phase following a TBI typically improve as thepathophysiology resolves. However, even mild neuroinflammation,

particularly when it occurs during development, has been shownto prime or sensitize the neuroimmune system to later inflamma-tory events. For instance, neonatal E. coli infection enhances thecognitive deficits and microglial responses to lipopolysaccharide-induced inflammation later in life (Bilbo et al., 2008, 2010). Fur-ther, TBI itself can act as a priming stimulus as microglial reactivityand behavioral depression induced by experimental inflammationis enhanced in mice with a history of mild traumatic brain injury(Fenn et al., 2014). A primary goal of the work presented herewas to determine whether an early life brain injury could similarlypotentiate the consequences of repeated high dose alcohol expo-sure in adulthood.

Fig. 5. Learning and memory are mildly impaired in alcohol-treated mice that sustained a juvenile TBI. A) All mice exhibit a similar latency to enter the dark side of thepassive avoidance chamber during training. B) However, the latency to re-enter the dark, shock-associated, side 24 h later was considerably longer, but was significantlyreduced in alcohol-treated mice that sustained a juvenile TBI. An asterisk (⁄) indicates statistically significant differences between the indicated groups (p < 0.05).

Fig. 6. Binge alcohol increases the number of ectopic doublecortin-positive cells in the hippocampus of mice that sustained a juvenile TBI. A) While the number of total DCX-positive cells did not differ in the granule cell layer, high doses of alcohol significantly increased the number of ectopic newborn neurons in the B) hilus and C) molecular celllayers of the hippocampus (white arrows). D) Representative images of the dentate gyrus, scale bar = 100 lm. An asterisk (⁄) indicates statistically significant differencesbetween the indicated groups (p < 0.05). ML = molecular layer, GCL = granule cell layer.

K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311 309

310 K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311

Neuroinflammatory responses are bidirectionally linked to alco-hol abuse as neuroinflammation can increase drinking behavior(Crews et al., 2011). Conversely, chronic exposure to high levelsof alcohol is neuroinflammatory as alcohol can activate glial cellsand promote the expression of cytokine and chemokine molecules(He and Crews, 2008). Although the mechanisms of alcohol-induced neuroinflammation are incompletely understood, alcoholdoes appear to signal in part through toll-like receptors (Alfonso-Loeches et al., 2010; Fernandez-Lizarbe et al., 2009). For instance,the long term behavioral and neurochemical effects of adolescentalcohol exposure require TLR4 (Montesinos et al., 2016).

Microglia isolated from the injured frontal lobes following thebinge-like administration of alcohol exhibited marked upregula-tion of mRNAs encoding the genes toll-like receptor 4 and IL-1b.Critically, this increase in expression of TLR4 and IL-1b onlyoccurred in mice that had been previously injured and treated withbinge-like levels of ethanol in adulthood. We have hypothesizedthat TBI sets up a vicious cycle wherein TBI during developmentprimes inflammatory responses to exhibit exaggerated inflamma-tory responses to alcohol, which in turn promotes increased drink-ing (Weil et al., 2016a). As an extension of this theory, theneurotoxic effects of juvenile TBI could also be exacerbated by highdoses of alcohol. The up-regulation of both TLR4, a potential medi-ator of alcohol-induced inflammation, and IL-1b, a key proinflam-matory cytokine, only in mice that had been injured and exposedto high levels of alcohol would tend to support this hypothesis. Itremains unspecified why we did not detect changes in geneexpression in microglia from sham-injured mice although this lar-gely agrees with our histological evidence including reducedmicroglial numbers, activated morphology and axonaldegeneration.

In the current study, we report exacerbation of the TBI-induceddamage to white matter tracts by binge-like levels of alcohol inaccordance with increased numbers and activational state ofmicroglia as well as exaggerated inflammatory gene expressionprofiles. Given that much of the tissue damage induced by chronicexposure to high levels of alcohol appears to require microglial-mediated inflammatory responses (Boyadjieva and Sarkar, 2010;Chastain and Sarkar, 2014; Yang et al., 2014), we conclude that atransient period of TBI-induced inflammation in juvenile develop-ment is sufficient to significantly exacerbate the neurotoxic effectsof binge alcohol in adulthood.

Interestingly we report a differential effect of TBI and alcoholexposure on microglial morphology, counts and gene expressionprofiles. Specifically, both TBI and alcohol exposure reduced thelength of microglial processes as assessed by Sholl analysis butonly microglia from mice that were exposed to alcohol and werepreviously injured were more numerous, had larger cell bodiesand upregulated TLR4 and IL1b. That microglial morphology doesnot fully track gene expression profiles is consistent with previousreports (McClain et al., 2011) and suggests that both alcohol expo-sure and TBI are independently associated with changes in micro-glial physiology and morphology but that can induce a fulleractivation when combined.

Moreover, exposure to binge-like levels of alcohol greatlyincreased the presence of immature doublecortin-positive neuronsin the non-neurogenic molecular layer and hilus of the dentategyrus but did so only in mice that had experienced TBI. Impor-tantly, the exposure to binge-like alcohol in the current studyoccurred more than six weeks after the TBI, thus the DCX-positive cells assessed at this time point were those cells that wereborn around the time of alcohol exposure rather than at the time ofinjury. Although the combination of injury and alcohol-exposureincreased the number of hilar DCX-positive cells, the total numberwas relatively small. Further there were no differences in the num-ber or localization of DCX-positive cells induced by injury alone

indicating that any acute effects of TBI on this cell populationhad resolved by adulthood.

CNS trauma often leads to an increase in the proliferation rate ofneuronal precursors, and this phenomenon has been explored as apotential replacement source for cells lost to trauma (Dash et al.,2001; Kernie and Parent, 2010). However, this strategy is compli-cated by reactive and aberrant neurogenesis that can occur afterinjury, seizure, or following withdrawal from heavy drinking whichcan render the brain more vulnerable to cognitive dysfunction,excitotoxicity, and additional seizures (Parent et al., 2006). Ectopicgranule cells that have been identified in the brains of epilepticanimals and humans exhibit abnormal morphology, connectivityand electrophysiological properties that contribute to hyperex-citability (Parent et al., 2006; Scharfman et al., 2000). Additionally,alcohol (and alcohol withdrawal) even in the absence of traumamodulates neurotransmission and generally biases towards excita-tory neurotransmission via changes in neuronal and glial neuro-transmitter metabolism, transport, and receptor expression(Adermark and Bowers, In Press; Dodd et al., 2000). Clinically,TBI patients at high risk for post-traumatic epilepsy are advisedto abstain from alcohol as drinking is positively correlated withthe occurrence of seizures (Centers for Disease Control andPrevention; Corrigan, 1995). The data collected here suggest thatan increased propensity for alcohol-induced ectopic neurogenesisin the previously injured brain may at least in part, underlie thisrelationship. Further, we cannot rule out the possibility that alco-hol or alcohol-withdrawal induced seizures may also haveoccurred and underlies the ectopic neurogenesis rather than theother way around.

The behavioral results reported here, namely relatively subtledeficits in simple associative learning and memory in the passiveavoidance paradigm, were conducted on animals experiencingalcohol withdrawal (retention trials occurred approximately 48 hafter their final dose of alcohol). Importantly, although there werestatistical differences between alcohol-and water-treated injuredmice there were not statistical differences between the shaminjured and TBI groups. In any case, we elected to test animals dur-ing withdrawal because we wanted to avoid any direct effects ofintoxication on behavioral outcomes. Additionally, it will beimportant in future investigations to determine whether the mildimpairments in learning and memory persist or improve beyondthe acute phase of alcohol withdrawal. Given the enhanced axonaldegeneration some degree of long-term functional deficit is likelyand may require more extensive behavioral characterization. Inany case, these data provide further compelling evidence thatbinge-like alcohol consumption negatively impacts functional out-comes in animals that were previously injured. Importantly, overthe 10-day treatment paradigm, mice were administered a singledaily high dose of alcohol every 24 h, thus in effect these animalsexperienced repeated cycles of intoxication and withdrawal overa short period of time. In addition, although we did not assessthe estrous cycle, the prolonged alcohol exposure paradigm didpresumably span at least two cycles. Finally, the paradigm we uti-lized produced high (but similar between sham-injured andinjured mice) blood alcohol concentrations. Such a high dose ofalcohol was utilized in order to evaluate whether neurotoxic doseswould produce worse outcomes in the previously injured brain,however, it remains unclear whether a longer exposure to moremoderate levels of alcohol would also be deleterious.

5. Conclusions

The human epidemiological literature has clearly demonstratedthat 1) the TBI patient population consists of a high proportion ofheavy and binge drinkers, 2) some percentage of TBI patients con-tinue to drink at high levels after TBI, and critically 3) that drinking

K. Karelina et al. / Brain, Behavior, and Immunity 60 (2017) 304–311 311

after TBI impairs functional and occupational recovery and puts thepatient at risk for serious complications including additional braininjuries and seizures. Here we begin to demonstrate the underlyingneurobiological mechanisms that underlie this phenomenonincluding alcohol-induced axonal degeneration, ectopic neurogen-esis, and enhanced neuroinflammation induced by alcohol in pre-viously injured mice. This vulnerability to alcohol-inducedhistopathology and functional impairment occurred despite a longdelay (more than 6 weeks) between the injury and the beginning ofbinge-alcohol administration. These data reinforce the importanceof targeting substance abuse in this population and are of evenmore importance given our previous observation that TBI can alsoincrease alcohol self-administration and reward.

Acknowledgments

This work was supported by an OSU Neurological Institute PilotAward and the National Institutes of Health (NS045758). Theauthors gratefully acknowledge Joseph Abraham, Katie Centner,Samuel Nicholson, Benjamin Sarac, and Lauren Martin for theirassistance and support on these experiments. We also thank Dr.Kristina Kigerl for guidance through the microglial extraction pro-tocol and Dr. Jacqueline Barker for assistance with the BEC assay.

References

Abràmoff, M.D., Magalhães, P.J., Ram, S.J., 2004. Image processing with ImageJ.Biophotonics Int. 11, 36–42.

Adermark, L., Bowers, M.S., In Press. Disentangling the Role of Astrocytes in AlcoholUse Disorder. Alcohol. Clin. Exp. Res.

Alfonso-Loeches, S., Pascual-Lucas, M., Blanco, A.M., Sanchez-Vera, I., Guerri, C.,2010. Pivotal role of TLR4 receptors in alcohol-induced neuroinflammation andbrain damage. J. Neurosci. 30, 8285–8295.

Berry, C., Ley, E.J., Margulies, D.R., Mirocha, J., Bukur, M., Malinoski, D., Salim, A.,2011. Correlating the blood alcohol concentration with outcome after traumaticbrain injury: too much is not a bad thing. Am. Surg. 77, 1416–1419.

Bertola, A., Mathews, S., Ki, S.H., Wang, H., Gao, B., 2013. Mouse model of chronicand binge ethanol feeding (the NIAAA model). Nat. Protoc. 8, 627–637.

Bilbo, S.D., Barrientos, R.M., Eads, A.S., Northcutt, A., Watkins, L.R., Rudy, J.W., Maier,S.F., 2008. Early-life infection leads to altered BDNF and IL-1beta mRNAexpression in rat hippocampus following learning in adulthood. Brain Behav.Immun. 22, 451–455.

Bilbo, S.D., Wieseler, J.L., Barrientos, R.M., Tsang, V., Watkins, L.R., Maier, S.F., 2010.Neonatal bacterial infection alters fever to live and simulated infections inadulthood. Psychoneuroendocrinology 35, 369–381.

Bjork, J.M., Grant, S.J., 2009. Does traumatic brain injury increase risk for substanceabuse? J. Neurotrauma 26, 1077–1082.

Bombardier, C.H., Temkin, N.R., Machamer, J., Dikmen, S.S., 2003. The natural historyof drinking and alcohol-related problems after traumatic brain injury. Arch.Phys. Med. Rehabil. 84, 185–191.

Boyadjieva, N.I., Sarkar, D.K., 2010. Role of microglia in ethanol’s apoptotic action onhypothalamic neuronal cells in primary cultures. Alcohol. Clin. Exp. Res. 34,1835–1842.

Brown, J.P., Couillard-Despres, S., Cooper-Kuhn, C.M., Winkler, J., Aigner, L., Kuhn, H.G., 2003. Transient expression of doublecortin during adult neurogenesis. J.Comp. Neurol. 467, 1–10.

Centers for Disease Control and Prevention, Facts about Concussion and BrainInjury: Where to get Help. <http://www.cdc.gov/headsup/pdfs/providers/facts_about_concussion_tbi-a.pdf>.

Chastain, L.G., Sarkar, D.K., 2014. Role of microglia in regulation of ethanolneurotoxic action. Int. Rev. Neurobiol. 118, 81–103.

Chen, C.M., Yi, H.-Y., Yoon, Y.-H., Dong, C., 2012. Alcohol use at time of injury andsurvival following traumatic brain injury: results from the National TraumaData Bank. J. Stud. Alcohol Drugs 73, 531–541.

Coronado, V.G., McGuire, L.C., Sarmiento, K., Bell, J., Lionbarger, M.R., Jones, C.D.,Geller, A.I., Khoury, N., Xu, L., 2012. Trends in Traumatic Brain Injury in the U.S.and the public health response: 1995–2009. J. Safety Res. 43, 299–307.

Corrigan, J.D., 1995. Substance abuse as a mediating factor in outcome fromtraumatic brain injury. Arch. Phys. Med. Rehabil. 76, 302–309.

Corrigan, J.D., Bogner, J., Mellick, D., Bushnik, T., Dams-O’Connor, K., Hammond, F.M., Hart, T., Kolakowsky-Hayner, S., 2013. Prior history of traumatic brain injuryamong persons in the traumatic brain injury model systems national database.Arch. Phys. Med. Rehabil. 94, 1940–1950.

Corrigan, J.D., Cuthbert, J.P., Harrison-Felix, C., Whiteneck, G.G., Bell, J.M., Miller, A.C., Coronado, V.G., Pretz, C.R., 2014. US population estimates of health and socialoutcomes 5 years after rehabilitation for traumatic brain injury. J. Head TraumaRehabil. 29, E1–E9.

Crews, F.T., Zou, J., Qin, L., 2011. Induction of innate immune genes in brain createthe neurobiology of addiction. Brain Behav. Immun. 25, S4–S12.

Dash, P.K., Mach, S.A., Moore, A.N., 2001. Enhanced neurogenesis in the rodenthippocampus following traumatic brain injury. J. Neurosci. Res. 63, 313–319.

Dodd, P.R., Beckmann, A.M., Davidson, M.S., Wilce, P.A., 2000. Glutamate-mediatedtransmission, alcohol, and alcoholism. Neurochem. Int. 37, 509–533.

Fenn, A.M., Gensel, J.C., Huang, Y., Popovich, P.G., Lifshitz, J., Godbout, J.P., 2014.Immune activation promotes depression 1 month after diffuse brain injury: arole for primed microglia. Biol. Psychiatry 76, 575–584.

Fernandez-Lizarbe, S., Pascual, M., Guerri, C., 2009. Critical role of TLR4 response inthe activation of microglia induced by ethanol. J. Immunol. 183, 4733–4744.

Harper, C., Dixon, G., Sheedy, D., Garrick, T., 2003. Neuropathological alterations inalcoholic brains. Studies arising from the New South Wales Tissue ResourceCentre. Prog. Neuropsychopharmacol. Biol. Psychiatry 27, 951–961.

Hayes, D.M., Deeny, M.A., Shaner, C.A., Nixon, K., 2013. Determining the thresholdfor alcohol-induced brain damage: new evidence with gliosis markers. Alcohol.Clin. Exp. Res. 37, 425–434.

He, J., Crews, F.T., 2008. Increased MCP-1 and microglia in various regions of thehuman alcoholic brain. Exp. Neurol. 210, 349–358.

Ikegami, Y., Goodenough, S., Inoue, Y., Dodd, P.R., Wilce, P.A., Matsumoto, I., 2003.Increased TUNEL positive cells in human alcoholic brains. Neurosci. Lett. 349,201–205.

Jorge, R.E., Starkstein, S.E., Arndt, S., Moser, D., Crespo-Facorro, B., Robinson, R.G.,2005. Alcohol misuse and mood disorders following traumatic brain injury.Arch. Gen. Psychiatry 62, 742–749.

Karelina, K., Sarac, B., Freeman, L.M., Gaier, K.R., Weil, Z.M., 2016. Traumatic braininjury and obesity induce persistent central insulin resistance. Eur. J. Neurosci.43, 1034–1043.

Kernie, S.G., Parent, J.M., 2010. Forebrain neurogenesis after focal Ischemic andtraumatic brain injury. Neurobiol. Dis. 37, 267–274.

Kreutzer, J.S., Harris, J., 1990. Model systems of treatment for alcohol abusefollowing traumatic brain injury. Brain Inj. 4, 1–5.

Langlois, J.A., Rutland-Brown, W., Wald, M.M., 2006. The epidemiology and impactof traumatic brain injury: a brief overview. J. Head Trauma Rehabil. 21, 375–378.

McClain, J.A., Morris, S.A., Deeny, M.A., Marshall, S.A., Hayes, D.M., Kiser, Z.M., Nixon,K., 2011. Adolescent binge alcohol exposure induces long-lasting partialactivation of microglia. Brain Behav. Immun. 25, S120–S128.

Montesinos, J., Pascual, M., Rodriguez-Arias, M., Minarro, J., Guerri, C., 2016.Involvement of TLR4 in the long-term epigenetic changes, rewarding andanxiety effects induced by intermittent ethanol treatment in adolescence. BrainBehav. Immun. 53, 159–171.

Opreanu, R.C., Kuhn, D., Basson, M.D., 2010. Influence of alcohol on mortality intraumatic brain injury. J. Am. Coll. Surg. 210, 997–1007.

Pandit, V., Patel, N., Rhee, P., Kulvatunyou, N., Aziz, H., Green, D.J., O’Keeffe, T.,Zangbar, B., Tang, A., Gries, L., Friese, R.S., Joseph, B., 2014. Effect of alcohol intraumatic brain injury: is it really protective? J. Surg. Res. 190, 634–639.

Parent, J.M., Elliott, R.C., Pleasure, S.J., Barbaro, N.M., Lowenstein, D.H., 2006.Aberrant seizure-induced neurogenesis in experimental temporal lobe epilepsy.Ann. Neurol. 59, 81–91.

Ponsford, J., Whelan-Goodinson, R., Bahar-Fuchs, A., 2007. Alcohol and drug usefollowing traumatic brain injury: a prospective study. Brain Inj. 21, 1385–1392.

Prencipe, L., Iaccheri, E., Manzati, C., 1987. Enzymic ethanol assay: a newcolorimetric method based on measurement of hydrogen peroxide. Clin.Chem. 33, 486–489.

Qin, L., Crews, F.T., 2012. Chronic ethanol increases systemic TLR3 agonist-inducedneuroinflammation and neurodegeneration. J. Neuroinflammation 9, 130.

Scharfman, H.E., Goodman, J.H., Sollas, A.L., 2000. Granule-like neurons at the hilar/CA3 border after status epilepticus and their synchrony with area CA3pyramidal cells: functional implications of seizure-induced neurogenesis. J.Neurosci. 20, 6144–6158.

Tagliaferri, F., Compagnone, C., Korsic, M., Servadei, F., Kraus, J., 2006. A systematicreview of brain injury epidemiology in Europe. Acta Neurochir. (Wien) 148,255–268. discussion 268.

Vaaramo, K., Puljula, J., Tetri, S., Juvela, S., Hillbom, M., 2014a. Head traumasustained under the influence of alcohol is a predictor for future traumatic braininjury: a long-term follow-up study. Eur. J. Neurol. 21, 293–298.

Vaaramo, K., Puljula, J., Tetri, S., Juvela, S., Hillbom, M., 2014b. Predictors of new-onset seizures: a 10-year follow-up of head trauma subjects with and withouttraumatic brain injury. J. Neurol. Neurosurg. Psychiatry 85, 598–602.

Vetreno, R.P., Crews, F.T., 2012. Adolescent binge drinking increases expression ofthe danger signal receptor agonist HMGB1 and Toll-like receptors in the adultprefrontal cortex. Neuroscience 226, 475–488.

Weil, Z.M., Corrigan, J.D., Karelina, K., 2016a. Alcohol abuse after traumatic braininjury: experimental and clinical evidence. Neurosci. Biobehav. Rev. 62, 89–99.

Weil, Z.M., Gaier, K.R., Karelina, K., 2014. Injury timing alters metabolic,inflammatory and functional outcomes following repeated mild traumaticbrain injury. Neurobiol. Dis. 70, 108–116.

Weil, Z.M., Karelina, K., Gaier, K.R., Corrigan, T.E., Corrigan, J.D., 2016b. Juveniletraumatic brain injury increases alcohol consumption and reward in femalemice. J. Neurotrauma 33, 895–903.

Yang, J.-Y., Xue, X., Tian, H., Wang, X.-X., Dong, Y.-X., Wang, F., Zhao, Y.-N., Yao, X.-C.,Cui, W., Wu, C.-F., 2014. Role of microglia in ethanol-inducedneurodegenerative disease: pathological and behavioral dysfunction atdifferent developmental stages. Pharmacol. Ther. 144, 321–337.