Epilepsy & Behavior 33 (2014) 138–143

Contents lists available at ScienceDirect

Epilepsy & Behavior

j ourna l homepage: www.e lsev ie r .com/ locate /yebeh

Intraperitoneal administration of docosahexaenoic acid for 14 daysincreases serum unesterified DHA and seizure latency in the maximalpentylenetetrazol model

Marc-Olivier Trépanier a,b,c, Joonbum Lim a,c, Terence K.Y. Lai a,c, Hye Jin Cho a,c, Anthony F. Domenichiello b,Chuck T. Chen b, Ameer Y. Taha a,c,1, Richard P. Bazinet b,c, W.M. Burnham a,c,⁎a Department of Pharmacology and Toxicology, Faculty of Medicine, University of Toronto, Toronto, ON M5S 1A8, Canadab Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, ON M5S 1A8, Canadac University of Toronto Epilepsy Research Program, Faculty of Medicine, University of Toronto, Toronto, ON M5S 1A8, Canada

⁎ Corresponding author at: Department of PharmacoloMedical Science Building, 1 King's College Circle, Toronto,

E-mail address: mac.burnham@utoronto.ca (W.M. Bur1 Present address: Brain Physiology and Metabolism

Aging, National Institutes of Health, 9000 Rockville Pike20892, USA.

http://dx.doi.org/10.1016/j.yebeh.2014.02.0201525-5050/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 23 October 2013Revised 4 January 2014Accepted 19 February 2014Available online xxxx

Keywords:Docosahexaenoic acidOmega-3 polyunsaturated fatty acidsSeizuresPentylenetetrazolRat

Docosahexaenoic acid (DHA) is an omega-3 polyunsaturated fatty acid (n−3 PUFA) which has been shown toraise seizure thresholds following acute administration in rats. The aims of the present experiment were the fol-lowing: 1) to test whether subchronic DHA administration raises seizure threshold in themaximal pentylenetet-razol (PTZ)model 24 h following the last injection and 2) to determinewhether the increase in seizure thresholdis correlated with an increase in serum and/or brain DHA.Animals received daily intraperitoneal (i.p.) injections of 50 mg/kg of DHA, DHA ethyl ester (DHAEE), or volume-matched vehicle (albumin/saline) for 14 days. On day 15, one subset of animals was seizure tested in the maxi-mal PTZ model (Experiment 1). In a separate (non-seizure tested) subset of animals, blood was collected, andbrains were excised following high-energy, head-focused microwave fixation. Lipid analysis was performed onserum and brain (Experiment 2). For data analysis, the DHA and DHA EE groups were combined since they didnot differ significantly from each other.In the maximal PTZ model, DHA significantly increased seizure latency by approximately 3-fold as comparedto vehicle-injected animals. This increase in seizure latencywas associatedwith an increase in serumunesterifiedDHA. Total brain DHA and brain unesterified DHA concentrations, however, did not differ significantly in thetreatment and control groups. An increase in serum unesterified DHA concentration reflecting increased flux ofDHA to the brain appears to explain changes in seizure threshold, independent of changes in brain DHAconcentrations.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

Epilepsy is a neurological disorder involving recurrent, spontaneousseizures. Approximately 1% of the population suffers from epilepsy atany given time, with approximately 3–4% of the population being af-fected at some point during their lives [1,2].

Antiepileptic drugs (AEDs) are the therapy of choice for epilepsy.However, there are approximately 30–40% of patients whose seizuresfail to respond to AEDs and continue to display seizures while on med-ication [3]. Moreover, many side effects are associated with AEDs, in-cluding nausea, drowsiness, and – in a rare subset of patients –

gy, MSB, University of Toronto,ON M5S 1A8, Canada.nham).Section, National Institute on, Room 1S-107, Bethesda, MD

hypersensitivity reactions [4,5]. Therefore, there is still a need for newtherapies for epilepsy.

One suggested therapy for epilepsy is dietary supplementation withomega-3 polyunsaturated fatty acids (n−3 PUFAs) [6]. n−3 PUFAs canbe derived from dietary sources such as seafood and are not known tohave any serious toxicity [7]. Docosahexaenoic acid (DHA) in particular,is an n−3 PUFA which is highly concentrated in the brain and plays arole in a number of brain functions including the regulation of gene ex-pression [8], the maintenance of membrane fluidity, signaling [9], andthe production of anti-inflammatory metabolites [10].

The n−3 PUFAs are known to have antiarrhythmic effects, which ap-pear to bemediated by their actions on voltage-dependent sodium chan-nels (VDSCs) [11,12]. In vitro work on hippocampal slices, for instance,has demonstrated that the n−3 PUFAs reduce sodium currents and in-crease VDSC inactivation time [13]. Other studies have shown that then−3 PUFAs increase the action potential firing threshold [14], reduce re-petitive action potential firing [14,15], and reduce hippocampal sharpwaves [16]. These effects on action potential firing are thought to be

139M.O. Trépanier et al. / Epilepsy & Behavior 33 (2014) 138–143

mediated by the unesterified form of the fatty acids, as ethyl esters areineffective [14]. It was recently shown, however, that both the freeand methyl ester forms of DHAwere effective at reducing hippocampalsharp wave firing [16].

It has been suggested that the n−3 PUFAsmight have anticonvulsantas well as antiarrhythmic effects. Several studies have been published onthe anticonvulsant properties of n−3 PUFAs in vivo. The first of these re-ported that intraperitoneal (i.p.) administration of 40 mg/kg of amixtureof linoleic acid (LA) and alpha-linolenic acid (ALA) in a 4:1 ratio for a pe-riod of 21 days resulted in seizure protection in 4 different seizuremodels [17]. Our own group was not able to reproduce this effect at40 mg/kg, but we did find an increase in seizure threshold at a higherdose of 200 mg/kg [18,19].

Alpha-linolenic acid is the precursor to longer chain n−3 PUFAs, in-cluding DHA. Studies investigating the specific effect of DHA haveshown DHA to increase seizure threshold in several in vivo studieswhen administered intravenously (i.v.) [20,21], subcutaneously (s.c.)[22,23], and intraperitoneally (i.p.) [21]. It has been previously demon-strated that acute administration of DHA results in increased seizure la-tency 1 h following injection, but this effect is lost 2 h postinjection [22].However, it has yet to be determined if multiple injections could in-crease the duration of efficacy of DHA on seizure latency.

The present studywas designed to investigate if subchronic injectioncan result in an elevation in seizure latency 24 h following the last injec-tion and whether this effect of DHA could be linked to elevations inunesterified DHA in serum and brain. Moreover, the efficacy of injectingthe ethyl ester form of DHA was also tested as it has yet to be testedin vivo. Docosahexaenoic acid was administered i.p. for a period of14 days. Animals were then seizure tested in the maximal PTZ model(Experiment 1) or sacrificed for measurement of DHA in serum andbrain (Experiment 2). The maximal PTZ test models tonic–clonic sei-zures in humans. Docosahexaenoic acid was administered both in anunesterified form (DHA) and in the form of an ethyl ester (DHA EE). Itwas hypothesized that the esterified form would be ineffective. It wasfound, however, that DHA EE was just as effective as DHA.

2. Materials and methods

2.1. Subjects

The present experiments were conducted in accordance with thestandards of the Canadian Council on Animal Care and were approvedby theAnimal Care Committee of the Faculty ofMedicine of theUniversityof Toronto. Sixty-day-old male Wistar rats were obtained from CharlesRiver (La Prairie, Qc) and were individually housed in a vivariummaintained on a 12 h light–dark cycle (lights on at 7 am) and at atemperature of 21 °C. Rat chow (2018 Teklad Global 18% Protein Ro-dent Diet; Teklad Global, Madison, WI) and water were available adlibitum. The rat chow contained (in g/kg diet) protein (189), fat(60), carbohydrate (554), fiber (38), ash (59), and moisture (100).The diet fat composition (in percent of total fatty acids) was palmi-tate (18.5%), stearate (2.8%), oleate (18.5%), linoleate (54.8%), andalpha-linolenate (5.6%) [23].

2.2. Drug preparation

Docosahexaenoic acid and docosahexaenoic acid ethyl ester wereobtained from Nu-Check Prep (Elysian, MN). Stock solutions were pre-pared by mixing 140 μl of DHA or DHA EE with 90 mg of bovineserumalbumin (BSA) perml of physiological saline. Thefinal concentra-tions of the stock solutionswere 128.8 mg/ml and 127.4 mg/ml for DHAand DHA EE, respectively (being equimolar in concentration). A vehiclesolution containing 90 mg of BSA dissolved in 1 ml of physiologicalsaline was given to control subjects. Stock solutions were sonicatedfor 5min. Stock solutions weremade on day 1 of injections andwere

stored at −80 °C between injections. Stock solutions were thawedprior to injections and were kept on ice during the injections.

Pentylenetetrazol was obtained from Sigma-Aldrich (Oakville, ON).The PTZ solution was prepared by dissolving 50 mg of PTZ per ml of0.9% saline. Pentylenetetrazol stock solution was made on day 15,prior to seizure testing, and was kept on ice between PTZ injections.

2.3. Experiment 1 — seizure testing

Two groups of 10 animals each received a dose of 50 mg/kg i.p. of ei-ther DHA or DHA EE for 14 consecutive days. Vehicle solution, volume-matched to theDHA group, was administered to a third group of controlanimals (n = 10).

On day 15, each animal received 105 mg/kg i.p. of PTZ. In a pilotstudy involving a separate group of subjects (n = 10), this dose of PTZhad been shown to reliably induce tonic–clonic convulsions. Followingthe PTZ injection, the subjects were placed in the open field and ob-served for 15 min. The latencies to the first myoclonic jerk and thefirst tonic–clonic seizure were scored by two independent observers.Following seizure testing, all subjects were immediately euthanizedwith a lethal intracardiac (i.c.) injection of 100 mg/kg of T-61 (Intervet,Kirkland, QC). As per animal care guidelines, if animals did not seizewithin 15 min, animals were euthanized by CO2 asphyxiation. Four an-imals (saline = 2, DHA = 1, and DHA EE = 1) did not have seizureswithin 15 min and were removed from analysis. No fatty acid analysiswas done in the seizure-tested subjects of Experiment 1.

2.4. Experiment 2 — blood and brain analysis

Two other groups of 19 non-seizure tested animals received a doseof 50 mg/kg i.p. of either DHA (n= 9) or DHA EE (n = 10) for 14 con-secutive days. Vehicle solution, volume-matched to the DHA group,wasadministered to a third group of control animals (n = 10).

On day 15, animals were placed under a heat lamp to generatevasodilation. Animals were then cannulated in the right tail vein with24-gauge (g) angiocath for blood sampling (BectonDickson,Mississauga,ON). Onemilliliter of bloodwas drawn. Bloodwas gently transferred intoa centrifuge tube and placed on ice. Blood was later spun for 5 min at4 °C, and serum was collected following centrifugation and stored at−80 °C.

Ten minutes following blood collection, animals were euthanizedwith high-energy, head-focused microwave fixation (13.5 kW for1.6 s; Cober Electronics Inc., Norwalk, CT) to stop brain lipidmetabolismand to prevent postmortem-related ischemic increases in unesterifiedfatty acid concentrations [24]. Heads were placed on dry ice to coolthem. Brains were then quickly removed from the skull and stored at−80 °C.

2.4.1. Folch extractionTotal lipidswere extracted from serumandbrain by Folch extraction

[25]. Total lipids were extracted from 200 μl serum in 2ml of methanol,4 ml of chloroform, and 1.6 ml of 0.88% KCl, followed by a second 4 mlchloroform wash. Unesterified (Nu-Chek Prep, Elysian, MN), triacyl-glyceride (TAG) (Nu-Chek Prep, Elysian, MN), and phosphatidyletha-nolamine (Avanti Polar Lipids, Inc., Alabaster, AL) heptadecanoic acidwere added as an internal standard. Total lipids were then dried downby nitrogen gas and reconstituted in 100 μl chloroform.

Brain samples were weighed prior to Folch extraction. Total lipidswere extracted from brain samples in 30 ml of chloroform and metha-nol (2:1 v/v) and 8ml of 0.88% KCl using a glass homogenizer. A second20 ml chloroform wash was performed. Heptadecanoic acid standard(10 nmol) was added to the total lipid extract. Total lipids were re-constituted into 1 ml chloroform.

Saline

DHA+DHA E

E0

100

200

300

400

Treatment

Late

ncy

to to

nic-

clon

ic s

eizu

res

(sec

)

*

Fig. 1. Latency of onset to tonic–clonic seizures following 14 days of daily saline albuminor DHA i.p. injections. Data for DHA and DHA EE were combined into one group (DHA +DHAEE) (n=18). Data are presented inmeans±SEM. Student's t-test found a significanteffect of DHA on latency of onset to tonic–clonic seizures compared with the saline albu-min control. *Significant difference as compared with the saline group as determined byStudent's t-test (p b 0.05).

0.0

0.1

0.2

0.3

0.4*

Une

ster

ified

DH

A s

erum

con

cent

ratio

n (µ

g/m

l)

Saline

DHA+DHA E

E

Treatment

Fig. 2. Serum unesterified DHA concentration (μg/ml) following 14 days of daily saline al-buminor DHA i.p. injections. Data are presented inmeans±SEM.DHA injections (n=19)significantly increased serum unesterified DHA concentrations compared with the salinealbumin control (n=10). *Significant difference as comparedwith the saline group as de-termined by Student's t-test (p b 0.05).

140 M.O. Trépanier et al. / Epilepsy & Behavior 33 (2014) 138–143

2.4.2. Thin layer chromatography (TLC)Thin layer chromatography (TLC) was used to separate serum

total lipids into free fatty acids (FFAs), phospholipids (PLs), andTAGs; TLC plates (G plate, EMD Chemicals, Gibbstown NJ) werewashed in chloroform methanol (2:1 v/v) and activated at 100 °Cfor 1 h. Total lipids (100 μl serum and 100 μl brain) were loaded ontothe plate and resolved in heptane, diethyl ether, and acetic acid(60:40:2 v/v) with neutral lipid standard lanes (Nu-Chek Prep, Elysian,MN) on both sides of the plate. Plates were sprayed with 0.1% (w/v)8-anilino-1-naphthalenesulfonic acid, and bands corresponding tototal PLs, FFAs, and TAGs were identified under ultraviolet light andscraped into new 15 ml glass screw cap tubes with Teflon lined caps.Hexane (2 ml) was added to the tubes. All three lipid fractionswere collected from serum samples, while only the unesterifiedfraction was collected from brain samples.

2.4.3. MethylationFrom brain total lipid extract, 50 μl was dried down and reconstituted

in 2 ml of hexane containing a known amount of unesterifiedheptadecanoic acid (Nu-Chek Prep, Elysian, MN). Brain total lipids (TLs)and unesterified fraction as well as serum FFA, PL, and TAG fractionswere then methylated with 2 ml 14% methanolic boron trifloride (BF3)(Sigma-Aldrich, Oakville, ON) for 1 h at 100 °C. Samples were cooled for10 min, and 2 ml of deionized water was added to the tubes. Tubeswere centrifuged for 10 min, and the hexane layer was extracted. Lipidswere reconstituted in 60 μl (serum — FFAs), 100 μl (serum — PLs andTAGs), 6 μl (brain— FFAs), and 500 μl (brain— TLs), and fatty acidmethylester (FAME) analysiswas performedby gas chromatography. It is impor-tant to note that since samples were methylated, it is not possible to dis-tinguish between unesterified DHA and DHA ethyl esters in our assay asboth lipids become DHA methyl ester following methylation.

2.4.4. Fatty acid methyl ester analysis by gas chromatographyFatty acid methyl esters were analyzed on a Varian-430 gas chro-

matograph (Varian, Lake Forest, CA) equipped with a Varian FactorFourcapillary column(VF-23ms; 30m×0.25mmi.d.×0.25 μmfilm thickness).One microliter of FAMEs was injected in splitless mode. The carrier gaswas helium, set to a constant flow rate of 0.7 ml/min. The injector anddetector ports were set at 250 °C. Fatty acid methyl esters were elutedusing a temperature program set initially at 50 °C for 2 min, followedby a ramp-up at 20 °C/min to 170 °C, and a hold at 170 °C for 1 minand an increase of 3 °C/min to 212 °C and a hold at 212 °C for 5 min.Peaks were confirmed by identifying the retention times of authenticFAME standards of known composition (Nu-Chek-Prep, Elysian, MN).Fatty acid concentrations (μg/ml of serum and nmol/g of brain) werecalculated by proportional comparisons of the gas chromatographypeak areas with that of the heptadecanoic acid internal standard.

2.5. Statistical analysis

Differences in the 3 groups (control, DHA, and DHA EE) for seizurelatency and DHA concentrations were initially analyzed by analysis ofvariance (ANOVA). Post-hoc analysis found no statistical differences be-tween the DHA and the DHA EE groups. The DHA and DHA EE groupswere subsequently combined to increase statistical power and namedDHA + DHA EE. Student's t-test was used to compare differences be-tween the saline and the DHA+ DHA EE groups.

3. Results

3.1. Experiment 1 — DHA and DHA EE increase seizure latency

In Experiment 1, animals received saline albumin (n = 10), DHA, orDHA EE (DHA + DHA EE, n = 20) for 14 consecutive days and weretested in the maximal PTZ model on day 15, 24 h after the last saline orDHA injection on day 14. As mentioned in the Materials and methods

section, 4 animals (saline = 2 and DHA + DHA EE = 2) were removedfrom the analysis due to lack of tonic–clonic seizures within 15 min ofPTZ injection. No significant differences in latencies to myoclonic jerks(53.1±6.0 vs. 145.2±46.0 s for saline andDHA+DHAEE, respectively)were found as determined by Student's t-test. Fig. 1 shows the averagelatencies (±standard error of the mean (SEM)) to tonic–clonic seizuresof control and DHA + DHA EE groups in the maximal PTZ model.

As illustrated by Fig. 1, the control group had a mean latency totonic–clonic seizures of 75.6 ± 9.8 s. In the DHA + DHA EE group, themean latency was increased to 286.0 ± 61.9 s. A Student's t-test founda significant difference between the two groups (p b 0.05).

3.2. Experiment 2 — DHA and DHA EE increase unesterified DHA serumconcentration but not PLs and TAGs

Experiment 2was designed to test the hypothesis that the increase inseizure latency observed in Experiment 1 was due to a specific increasein the unesterified serum and brain DHA concentrations. Serum and

141M.O. Trépanier et al. / Epilepsy & Behavior 33 (2014) 138–143

brain DHA concentrations were measured in several pools following 14days of daily i.p. injections.

Fig. 2 illustrates the unesterified DHA concentrations in serumfollowing 14 days of saline albumin or DHA injections. Daily DHA orDHA EE (DHA + DHA EE) injection increased unesterified DHA in theserum to 0.33 ± 0.03, as compared with 0.21 ± 0.03 μg/ml in the con-trol group. Analysis by Student's t-test revealed a significant differencebetween the DHA + DHA EE group and the control group (p b 0.05,Fig. 2).

Fourteen days of daily injections of DHA or DHA EE did not increaseserum DHA concentration in either the PL or the TAG fraction (data notshown). Student's t-test found no significant differences on the serumconcentrations in both lipid fractions between the control andDHA + DHA EE groups (p N 0.05).

3.3. Experiment 2 — DHA and DHA EE do not modify total and unesterifiedbrain DHA

Fig. 3A presents total brain DHA concentrations following 14 days ofsaline albumin or DHA injections. As illustrated in the figure, DHA andDHA EE injections had no effect on total brain DHA concentrations. AStudent's t-test found no significant differences between the controland DHA + DHA EE groups (p N 0.05).

Saline

DHA+DHA E

E

Treatment

Saline

DHA+DHA E

E

Treatment

8000

6000

4000

2000

0

Tot

al B

rain

DH

A (

nmol

/g)

0.0

0.5

1.0

1.5

Une

ster

ified

DH

A b

rain

con

cent

ratio

n (n

mol

/g)

A

B

Fig. 3.A. Brain total DHA concentration (nmol/g) following 14 days of daily saline albuminor DHA i.p. injections. Data are presented in means ± SEM. No significant differences intotal brain DHA concentrations were found between the saline albumin (n = 10) andDHA + DHA EE (n = 19) groups by Student's t-test (p N 0.05). B. Brain unesterifiedDHA concentration (nmol/g) following 14 days of daily saline albumin or DHA i.p. injec-tions. Data are presented in means ± SEM. No significant differences in total brain DHAconcentrations were found between saline albumin (n = 8) and DHA + DHA EE (n =18) as evaluated by Student's t-test (p N 0.05). Three sampleswere removed from analysis(saline = 2 and DHA + DHA EE = 1) because values fell below the GC detection limit.

The data related to the unesterified DHA concentrations aftertwo weeks of treatment are presented in Fig. 3B. Mean (±SEM) brainunesterified DHA concentrations were 0.46 ± 0.15 and 0.94 ±0.25 nmol/g of brain for control and DHA + DHA EE, respectively(three animals, saline = 2 and DHA + DHA EE = 1, had unesterifiedDHA concentration below the detection limit of the gas-chromatographand were removed from analysis). No significant differences betweentreatment groups were found by Student's t-test (p N 0.05).

4. Discussion

The present study has demonstrated that the daily administration of50 mg/kg of DHA for 14 days increases seizure threshold in themaximalPTZmodel. This increase in seizure thresholdwas correlatedwith an in-crease in DHA concentration in the unesterified fatty acid pool in serum.Contrary to our hypothesis, however, there was no significant increasein brain unesterified or total DHA concentrations, although there wasa “trend” toward elevation in the unesterified pool. Our data are in gen-eral agreement with the previous finding of Yehuda et al. who demon-strated that the administration for 21 days of 40 mg/kg of a mixture ofLA and ALA in a 4:1 ratio increased seizure threshold in four seizuremodels [17,26]. While our group was not able to replicate this effect inthe PTZ model at 40 mg/kg, [18], we did find threshold rises at thehigher concentration of 200 mg/kg per day [19].

It was argued in previous studies of ALA and LA that their anticon-vulsant effects were due to increases in DHA synthesis [27] or by in-creased mobilization of n−3 PUFAs from the liver to the brain [28].This is in agreement with the results of the present study in whichDHA was administered directly. The present results are also in agree-ment with the results of past studies showing anticonvulsant effects ofDHA directly administered at several different doses and by several dif-ferent routes [20–23].

Since an anticonvulsant effect of DHA was observed following14 days of i.p. administration, we hypothesized that the anticonvulsanteffects would be related with an increase in serum and brain DHA. Invitro studies suggest that the anticonvulsant effect of DHA is mediatedby unesterified DHA [14,16], although free DHA methyl ester has alsobeen shown to reduce hippocampal sharp waves [16].

The present study did find a significant increase in serum unes-terifiedDHA in the experimental animals, as comparedwith the animalsreceiving saline solution. This increase was lipid-pool specific, as the PLand TAG pool were unaffected by DHA administration. We previouslyhave demonstrated that acute s.c. administration of DHA increased sei-zure latency 1 h following injection [22,23], and this effect was lost 2 hafter the injection [22]. At 1 h post-s.c. DHA, serum unesterified DHAwas elevated. Contrary to the loss of effect observed 2 h post-acute s.c.injection [22], in this study, we observed increased seizure latency 24h following the last i.p. DHA injection. This effect was correlated withan increase in serum unesterified DHA.

The current proposed mechanism for fatty acid entry into the brainis passive diffusion from the circulating unesterified fatty acids[29–31]. In this study, this specific increase in the serum unesterifiedDHA, however, was not correlated with an increase in brain total DHA.This result is not surprising, as the unesterified DHA concentration inthe brain is approximately 9000-fold lower compared with the totalbrain DHA (Fig. 3).

Contrary to our hypothesis, however, no increase in unesterifiedDHA was observed in the brain. One possible explanation is that brainunesterified DHA levels were elevated, but our analytical techniqueswere not sensitive enough to detect the increase. Brain unesterifiedDHA concentrations are in the low nanomolar range, and, at these con-centrations, small differences between experimental and control levelsmight have been obscured because of random variability inmeasurement.

A second possibility may be that while brain unesterified DHA didnot increase, DHA turnover from the phospholipid could have been

142 M.O. Trépanier et al. / Epilepsy & Behavior 33 (2014) 138–143

increased. It has previously been shown that chronic administrationof valproate, a broad-spectrum anticonvulsant, reduces arachidonicacid turnover without affecting the unesterified concentrations [32].Conversely, it is possible that chronic administration of DHA in our ex-periment resulted in an increase in the turnover of DHA from the phos-pholipid pool without altering its unesterified concentration.

In the current study, increases in seizure latency were achievedfollowing 14 days of chronic administration of i.p. DHA. Anticonvulsanteffects have been similarly reported in a number of other studies involv-ing short-term parenteral DHA administration in humans or rats[20–23]. In contrast, it takes several months to raise seizure thresholdin rats when DHA is administered via the oral route [33–35]. This dis-crepancy may be due to the fact that fatty acids are packaged into chy-lomicrons when ingested orally [36]. Fatty acids packaged inlipoproteins are not available for brain uptake [29–31] and are packagedinto the adipose tissue [37,38]. From adipose, fatty acids are cleaved andreleased into the serum unesterified pool. However, since the half life ofn−3 PUFAs in rat adipose tissue is approximately 20 days [39], it maybe that a long duration of administration is required in order to increaseadipose n−3 concentration and, consequently, serum unesterifiedpool.

This is thefirst study, to our knowledge, to test effects of direct injec-tion of the ethyl ester form of DHA on seizure latency. It has previouslybeen demonstrated that oral administration of the ethyl form of DHAdid not convey seizure protection in 4 different models despite in-creases in total serum DHA [40]. Unesterified serum DHA, however,was not measured and perhaps may not have been elevated following1 month of oral DHA ethyl ester feeding.

If similar pharmacokinetics apply in humans, it may explain themixed results reported in human trials involving dietary supplementa-tionwith n−3 PUFAs. In clinical trials involving n−3 PUFA supplemen-tation, threemonths of administration of n−3 PUFAs has failed to showanticonvulsant effects [41–44], whereas 6 months of administration ofn−3 PUFAs has been shown to be anticonvulsant [45,46].

In conclusion, the present study has demonstrated that daily i.p.administration of DHA for 2 weeks increases seizure latency 24 h fol-lowing the last injection in the maximal PTZ seizure model, reflectingan increase in seizure threshold. This increase in seizure threshold wasassociated with an increase in serum DHA in the unesterified fattyacid pool but no significant increase in brain DHA. This study supportsthe concept that DHA could possibly be used as a therapy for epilepsy.It also appears to highlight the importance of the unesterified pool inthe blood as a biomarker for the anticonvulsant properties of DHA.

In the future, clinical trials should attempt to use longer periods ofadministration in order to assure an increase in unesterified DHA inserum and should attempt to relate clinical efficacy to concentrationchanges in the unesterified pool instead of the phospholipid or totallipid pool [46,47].

Conflict of interest statement

The authors declare no conflict of interest.

Acknowledgment

This study was funded, in part, by a grant from the CanadianInstitutes of Health Research (grant # mop-89364) and the BahenChair in Epilepsy grant toW.M.B. and, in part, by a grant from the Natu-ral Health Sciences and Research Council of Canada (grant #482597) toR.P.B., who holds a Canada Research Chair in brain lipid metabolism.

References

[1] Guberman A, Bruni J. Essentials of clinical epilepsy. 2nd ed. Boston: Butterworth-Heinemann; 1999.

[2] McHugh JC, Delanty N. Epidemiology and classification of epilepsy: gender compar-isons. Int Rev Neurobiol 2008;83:11–26.

[3] Burnham WM. Antiseizure drugs. In: Kalant H, Grant DM, Mitchell J, editors. Princi-ples of medical pharmacology. 7th ed. Toronto: Saunders; 2007. p. 223–35.

[4] Perucca P, Gilliam FG. Adverse effects of antiepileptic drugs. Lancet Neurol 2012;11:792–802.

[5] KrivoyN, TaerM, NeumanMG. Antiepileptic drug-induced hypersensitivity syndromereactions. Curr Drug Saf 2006;1:289–99.

[6] Taha AY, BurnhamWM, Auvin S. Polyunsaturated fatty acids and epilepsy. Epilepsia2010;51:1348–58.

[7] Russo GL. Dietary n−6 and n−3 polyunsaturated fatty acids: from biochemistry toclinical implications in cardiovascular prevention. Biochem Pharmacol 2009;77:937–46.

[8] Yavin E. Versatile roles of docosahexaenoic acid in the prenatal brain: from pro- andanti-oxidant features to regulation of gene expression. Prostaglandins Leukot EssentFatty Acids 2006;75:203–11.

[9] Jones CR, Arai T, Rapoport SI. Evidence for the involvement of docosahexaenoic acidin cholinergic stimulated signal transduction at the synapse. Neurochem Res1997;22:663–70.

[10] Bazan NG, Eady TN, Khoutorova L, Atkins KD, Hong S, Lu Y, et al. Novel aspirin-triggered neuroprotectin D1 attenuates cerebral ischemic injury after experimentalstroke. Exp Neurol 2012;236:122–30.

[11] Kang JX, Leaf A. Evidence that free polyunsaturated fatty acidsmodifyNa+ channelsby directly binding to the channel proteins. Proc Natl Acad Sci U S A 1996;93:3542–6.

[12] Kang JX, Xiao YF, Leaf A. Free, long-chain, polyunsaturated fatty acids reduce mem-brane electrical excitability in neonatal rat cardiac myocytes. Proc Natl Acad Sci U S A1995;92:3997–4001.

[13] Vreugdenhil M, Bruehl C, Voskuyl RA, Kang JX, Leaf A, WadmanWJ. Polyunsaturatedfatty acids modulate sodium and calcium currents in CA1 neurons. Proc Natl AcadSci U S A 1996;93:12559–63.

[14] Xiao Y, Li X. Polyunsaturated fatty acids modify mouse hippocampal neuronal excit-ability during excitotoxic or convulsant stimulation. Brain Res 1999;846:112–21.

[15] Young C, Gean PW, Chiou LC, Shen YZ. Docosahexaenoic acid inhibits synaptic trans-mission and epileptiform activity in the rat hippocampus. Synapse 2000;37:90–4.

[16] Taha AY, Zahid T, Epps T, Trepanier MO, Burnham WM, Bazinet RP, Zhang L. Selec-tive reduction of excitatory hippocampal sharp waves by docosahexaenoic acidand its methyl ester analog ex-vivo. Brain Res 2013;1537:9–17.

[17] Yehuda S, Carasso RL, Mostofsky DI. Essential fatty acid preparation (SR-3) raises theseizure threshold in rats. Eur J Pharmacol 1994;254:193–8.

[18] Taha AY, Baghiu BM, Lui R, Nylen K,MaDW, BurnhamWM. Lack of benefit of linoleicand alpha-linolenic polyunsaturated fatty acids on seizure latency, duration, severityor incidence in rats. Epilepsy Res 2006;71:40–6.

[19] Taha AY, Filo E, Ma DW, McIntyre Burnham W. Dose-dependent anticonvulsant ef-fects of linoleic and alpha-linolenic polyunsaturated fatty acids on pentylenetetrazolinduced seizures in rats. Epilepsia 2009;50:72–82.

[20] Voskuyl RA, Vreugdenhil M, Kang JX, Leaf A. Anticonvulsant effect of polyunsaturatedfatty acids in rats, using the cortical stimulation model. Eur J Pharmacol 1998;341:145–52.

[21] Musto AE, Gjorstrup P, Bazan NG. The omega-3 fatty acid-derived neuroprotectin D1limits hippocampal hyperexcitability and seizure susceptibility in kindlingepileptogenesis. Epilepsia 2011;52:1601–8.

[22] Taha AY, Jeffrey MA, Taha NM, Bala S, Burnham WM. Acute administration ofdocosahexaenoic acid increases resistance to pentylenetetrazol-induced seizuresin rats. Epilepsy Behav 2010;17:336–43.

[23] Trepanier MO, Taha AY, Mantha RL, Ciobanu FA, Zeng QH, Tchkhartichvili GM, et al.Increases in seizure latencies induced by subcutaneous docosahexaenoic acid arelost at higher doses. Epilepsy Res 2012;99:225–32.

[24] Bazan Jr NG. Effects of ischemia and electroconvulsive shock on free fatty acid poolin the brain. Biochim Biophys Acta 1970;218:1–10.

[25] Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purifica-tion of total lipides from animal tissues. J Biol Chem 1957;226:497–509.

[26] Rabinovitz S, Mostofsky DI, Yehuda S. Anticonvulsant efficiency, behavioral perfor-mance and cortisol levels: a comparison of carbamazepine (CBZ) and a fatty acidcompound (SR-3). Psychoneuroendocrinology 2004;29:113–24.

[27] Yehuda S, Brandys Y, Blumenfeld A, Mostofsky DI. Essential fatty acid preparationreduces cholesterol and fatty acids in rat cortex. Int J Neurosci 1996;86:249–56.

[28] Taha AY, Alizadeh S, Zeng QH, Filo E, McPherson JP, Burnham WM. Assessing themetabolic and toxic effects of anticonvulsant doses of polyunsaturated fatty acidson the liver in rats. J Toxicol Environ Health A 2009;72:1191–200.

[29] Song BJ, Elbert A, Rahman T, Orr SK, Chen CT, Febbraio M, et al. Genetic ablation ofCD36 does not alter mouse brain polyunsaturated fatty acid concentrations. Lipids2010;45:291–9.

[30] Chen CT, Ma DW, Kim JH, Mount HT, Bazinet RP. The low density lipoprotein recep-tor is not necessary for maintaining mouse brain polyunsaturated fatty acid concen-trations. J Lipid Res 2008;49:147–52.

[31] Ouellet M, Emond V, Chen CT, Julien C, Bourasset F, Oddo S, et al. Diffusion ofdocosahexaenoic and eicosapentaenoic acids through the blood-brain barrier: anin situ cerebral perfusion study. Neurochem Int 2009;55:476–82.

[32] Chang MC, Contreras MA, Rosenberger TA, Rintala JJ, Bell JM, Rapoport SI. Chronicvalproate treatment decreases the in vivo turnover of arachidonic acid in brain phos-pholipids: a possible common effect of mood stabilizers. J Neurochem 2001;77:796–803.

[33] Taha AY, Trepanier MO, Ciobanu FA, Taha NM, Ahmed M, Zeng Q, et al. A minimumof 3 months of dietary fish oil supplementation is required to raise amygdaloidafterdischarge seizure thresholds in rats — implications for treating complex partialseizures. Epilepsy Behav 2013;27:49–58.

143M.O. Trépanier et al. / Epilepsy & Behavior 33 (2014) 138–143

[34] Gilby KL, Jans J, McIntyre DC. Chronic omega-3 supplementation in seizure-proneversus seizure-resistant rat strains: a cautionary tale. Neuroscience 2009;163:750–8.

[35] Bandero CR, SalvadoriMG, Gomes AT, Dal Ri NM, Furian AF, OliveiraMS, et al. Fish oilattenuates methylmalonate-induced seizures. Epilepsy Res 2013;105:69–76.

[36] Mansbach CM, Siddiqi SA. The biogenesis of chylomicrons. Annu Rev Physiol 2010;72:315–33.

[37] Tso P, Balint JA. Formation and transport of chylomicrons by enterocytes to thelymphatics. Am J Physiol 1986;250:G715–26.

[38] Mayes PA, BothamKM. Lipid Transport & Storage. In: Murray RK, Granner DK, MayesPA, Rodwell VW, editors. Harper's Illustrated Biochemistry (26th ed). New York:McGraw-Hill; 2003. p. 205–18.

[39] Pihl A, Block K, Anker HS. The rates of synthesis of fatty acids and cholesterol in theadult rat studied with the aid of labeled acetic acid. J Biol Chem 1950;183:10.

[40] Willis S, Samala R, Rosenberger TA, Borges K. Eicosapentaenoic and docosahexaenoicacids are not anticonvulsant or neuroprotective in acute mouse seizure models.Epilepsia 2009;50:138–42.

[41] Yuen AW, Flugel D, Poepel A, Bell GS, Peacock JL, Sander JW. Non-randomized opentrial of eicosapentaenoic acid (EPA), an omega-3 fatty acid, in ten peoplewith chronicepilepsy. Epilepsy Behav 2012;23:370–2.

[42] Yuen AW, Sander JW, Flugel D, Patsalos PN, Browning L, Bell GS, et al. Erythrocyteand plasma fatty acid profiles in patients with epilepsy: does carbamazepine affectomega-3 fatty acid concentrations? Epilepsy Behav 2008;12:317–23.

[43] Bromfield E, Dworetzky B, Hurwitz S, Eluri Z, Lane L, Replansky S, et al. A randomizedtrial of polyunsaturated fatty acids for refractory epilepsy. Epilepsy Behav 2008;12:187–90.

[44] DeGiorgio CM, Miller P, Meymandi S, Gornbein JA. n−3 fatty acids (fish oil) forepilepsy, cardiac risk factors, and risk of SUDEP: clues from a pilot, double-blind,exploratory study. Epilepsy Behav 2008;13:681–4.

[45] Schlanger S, Shinitzky M, Yam D. Diet enriched with omega-3 fatty acids alleviatesconvulsion symptoms in epilepsy patients. Epilepsia 2002;43:103–4.

[46] Al Khayat HA, AwadallaMM, AlWakad A,Marzook ZA. Polyunsaturated fatty acids inchildrenwith idiopathic intractable epilepsy: serum levels and therapeutic response.J Pediatr Neurol 2010;8:175–85.

[47] Yuen AW, Sander JW, Fluegel D, Patsalos PN, Bell GS, Johnson T, et al. Omega-3 fattyacid supplementation in patients with chronic epilepsy: a randomized trial. EpilepsyBehav 2005;7:253–8.