A

(mfittcto©

K

1

ottbGMaadtta

sf

0d

Journal of Neuroscience Methods 169 (2008) 135–140

Low-temperature improved-throughput method for analysis of brain fattyacids and assessment of their post-mortem stability

Thomas Fraser ∗, Hannah Tayler, Seth LoveDementia Research Group, Institute of Clinical Neurosciences, Clinical Science at North Bristol, University of Bristol,

Frenchay Hospital, Frenchay, Bristol BS16 1LE, UK

Received 3 September 2007; received in revised form 29 November 2007; accepted 3 December 2007

bstract

Deficiency of docosahexaenoic acid (DHA) and other omega-3 (�3) fatty acids may constitute an alterable risk factor for Alzheimer’s diseaseAD). Mechanisms of potential involvement of DHA in the disease process have been postulated primarily from studies in vitro and in mouseodels of AD. Information on the fatty acid profile of the brain in AD itself is limited and in some respects contradictory. Interpretation of thendings is complicated by the diversity of methods used in previous studies and a lack of information as to the effect of post-mortem delay on

he results. Here we report the development of a simple and highly reproducible method that enables relatively high-throughput measurement ofhe fatty acid composition in samples of brain tissue and using this method we have demonstrated that there is no significant change in fatty acid

omposition under conditions designed to model post-mortem delay of up to 3 days at 4 ◦C (or even at room temperature). The development ofhis method and the observation that delay of up to 3 days has no effect on fatty acid content will facilitate further studies of fatty acid compositionn large cohorts of post-mortem brains. 2007 Elsevier B.V. All rights reserved.

Lipid

rdiearnVlD2cfLs

eywords: Docosahexaenoic acid; Methylation; Saponification; Post-mortem;

. Introduction

Evidence that brain lipid metabolism may influence the devel-pment of several neurological diseases, including AD, includeshe observations that (i) the gene encoding the cholesterolransporter apolipoprotein E type 4 is a major risk factor foroth sporadic and familial AD, (ii) epidemiological (Barberger-ateau et al., 2002; Conquer et al., 2000; Larrieu et al., 2004;orris et al., 2003; Tully et al., 2003; Young and Conquer, 2005)

nd experimental studies have indicated that increased �3 fattycid consumption decreases the risk of developing a number ofiseases including AD (reviewed in Assisi et al., 2006) and (iii)he modern human diet is deficient in �3 fatty acids comparedo the paleolithic diet on which humans evolved (Broadhurst etl., 2002; Muskiet et al., 2004).

Dietary ingestion of the brain-essential 22 carbon �3 polyun-aturated fatty acid, docosahexaenoic acid (DHA; 22:6n-3),ound predominantly in marine, oily fish and algae, has been

∗ Corresponding author. Tel.: +44 117 9701212x3068; fax: +44 117 9573955.E-mail address: t.c.m.fraser@bristol.ac.uk (T. Fraser).

2t(tep2

165-0270/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.jneumeth.2007.12.001

epeatedly implicated as a potential modifier of the risk ofevelopment of AD. This fatty acid, which is concentratedn synaptic membranes (Breckenridge et al., 1972), is consid-red critical for normal learning and memory (Hashimoto etl., 2002; Moriguchi et al., 2000; Salem et al., 2001) and neu-odevelopment (Lauritzen et al., 2001). DHA is necessary forormal membrane fluidity and function (Farkas et al., 2000;alentine and Valentine, 2004). Dietary manipulation of DHA

evels in a transgenic mouse model of AD has shown thatHA protects against dendritic degeneration (Calon et al.,004). DHA deficiency is associated with oxidative stress andaspase-mediated cleavage of actin, which may be responsibleor damage to dendritic scaffold proteins (Calon et al., 2004).ow brain DHA levels cause a decline in NMDA receptor den-ity (Calon et al., 2005). DHA has anti-oxidant activity (Bazan,005; Calon et al., 2004; Cole et al., 2004) and is a poten-ial stimulator of cerebral activities of catalase and glutathioneHossain et al., 1999). DHA lowers cerebral A� load and poten-

ial downstream toxicity in an aged mouse model of AD (Limt al., 2005) and, accompanied by the biosynthesis of, neuro-rotectin D1, lowers neuronal secretion of A� (Lukiw et al.,005).

1 scienc

AataanrntsaBo(

pDtdaommrtp

2

2

tEvwS4ss

2

f(ta–

2

am6z

abwwwfta(wfwlc

2

(e

2

ewuwhtuvrtwrwa

2

pat(

2

acUK) and a Shimadzu GC-14B series gas chromatograph with

36 T. Fraser et al. / Journal of Neuro

DHA released from membrane lipid stores by phospholipase2 provides substrate for the biosynthesis of D class resolvins,

nd neuroprotectins including neuroprotectin D1. These bioac-ive docosanoids inhibit the synthesis of pro-inflammatoryrachidonic acid (20:4 n-6) derived prostaglandins leukotrienesnd thromboxanes and are implicated in the prevention ofeuroinflammation associated with acute neural trauma and neu-odegenerative diseases (Farooqui et al., 2007). DHA metaboliteeuroprotectin D1 also inhibits A�42-induced apoptosis of cul-ured human neural cells (Lukiw et al., 2005), is a potenttimulant for resolution of inflammation (Schwab et al., 2007)nd a key component of survival signalling in AD (Lukiw andazan, 2006). Neuroprostanes, free radical-mediated derivativesf DHA, are increased in the brain and cerebrospinal fluid in ADMontine et al., 2004; Reich et al., 2001).

Although much information is therefore available about theotential involvement of DHA in AD, it is presently unclear howHA levels are affected in this disease, as the fatty acid composi-

ion of brain tissue has been analysed in only a few studies, usingifferent methods and based on small numbers of samples withrange of post-mortem delays. Here we report the developmentf a method that enables relatively rapid and accurate measure-ent of cerebral fatty acid composition. We have, in addition,odelled post-mortem delay of various durations at 4 ◦C and at

oom temperature and shown that storage of brain tissue for upo 72 h, even at room temperature, has no effect on the fatty acidrofile of frontal cortex as determined by this method.

. Materials and methods

.1. Materials

Brain tissue was obtained from the South West Demen-ia Brain Bank, University of Bristol, with local Researchthics Committee approval. Two ml screw-top sealable glassials purchased from Fisher Scientific (Loughborough, UK)ere used during experimental procedures unless indicated.olvents and butylated hydroxytoluene (BHT; 2,6-di-tert-butyl--methylphenol) which was added to chloroform–methanololutions at a concentration of 100 mg/l were also from thisupplier.

.2. Lipid extraction

Samples of anterior frontal cortex (Brodmann area 10)rom four normal human brains with short post-mortem delays4–10 h) were used for lipid extraction by a procedure based onhe Folch method. Centrifugation steps were conducted for 5 mint 2000 × g. Four replicates – one sample from each normal brainwere used for analysis of every experimental condition.

.2.1. Mechanical homogenisationBrain tissue (50 mg of frontal cortex) in 350 �l of methanol

t 0 ◦C was homogenised in a Bertin Technologies Precellys 24echanical homogeniser (Stretton Scientific, Stretton, UK) at

000 rpm for 2 s × 30 s with 15–25 2.3 mm Biospec Productsirconia–silica beads (Stratech Scientific, Newmarket, UK) in

fl2i(

e Methods 169 (2008) 135–140

screw cap 2 ml microcentrifuge tube that had been washedy vortexing with 1 ml of chloroform three times. The sampleas cooled on ice after homogenisation. Chloroform (400 �l)as added and the solution briefly vortexed. The homogenateas removed and the beads/microcentrifuge tube washed with a

urther 150 �l of chloroform:methanol 2:1 (v:v). The two solu-ions were combined and spun. The supernatant was removednd the pellet vortexed in 200 �l of chloroform:methanol 1:1v:v). The resuspended pellet was re-spun. The two supernatantsere combined and after addition of a further 400 �l of chloro-

orm and 325 �l of 0.9% (w/v) potassium chloride the solutionas vortexed and spun. The lipid fraction obtained from the

ower chloroform phase was used to determine the fatty acidomposition of neural tissues.

.2.2. Manual homogenisationBrain tissue was homogenized in chloroform:methanol 2:1

v:v) using a Potter-Elvehjem homogeniser and the lipids werextracted by the Folch method (Folch et al., 1957).

.3. Saponification

The chloroform phase was removed under N2 and 200 �l ofthanolic KOH (1 M) was added to the lipid residue. The solutionas capped and left overnight at room temperature or cappednder nitrogen and kept at 80 ◦C for 1 h. Distilled water (1 ml)as added to the saponification mixture along with 0.7 ml ofexane:diethylether 1:1 (v:v) and the biphasic solution was vor-exed and spun for 5 min at 2000 × g. The cholesterol containingpper phase was discarded. The addition of hexane:diethylether,ortexing, centrifugation and removal of upper phase steps wereepeated three times. The pH of the lower phase was adjustedo pH 3 with 5 M hydrochloric acid. 600 �l of diethyl etheras added followed by vortexing, centrifugation (as above) and

emoval of the upper phase repeated twice and the lower phaseas discarded. The two upper phases (containing the free fatty

cids) were combined and dried under N2.

.4. Methylation

Methyl esters of free fatty acids were prepared as describedreviously with a solution of 2.5% concentrated H2SO4 innhydrous methanol either capped and left overnight at roomemperature or capped under nitrogen and kept at 80 ◦C for 1 hFraser et al., 2004).

.5. Analysis of fatty acid methyl esters

Fatty acid methyl esters were separated and quantified using30 m × 0.25 mm fused silica DB-23 column J and W Scientificapillary column supplied by Fisher Scientific (Loughborough,

ame ionization detection as described previously (Qi et al.,004). Fatty acids methyl esters were identified by compar-son of retention time with those of authenticated standardsSigma–Aldrich, UK).

T. Fraser et al. / Journal of Neuroscience Methods 169 (2008) 135–140 137

Table 1Relevance of reaction temperature, capping under nitrogen and mechanical vs. manual homogenisation to analysis of fatty acid content

Saponification/methylation RT High temperature

Homogenisation Mechanical Manual Mechanical

Capped under nitrogen Not capped Capped Not capped Capped

Fatty acid Level of fatty acid ± S.E. (�mol/g wet weight)

16:0 14.79 ± 0.60 15.36 ± 0.63 15.55 ± 0.9 15.51 ± 0.8318:0 16.03 ± 0.98 16.21 ± 0.86 16.46 ± 0.84 16.19 ± 1.3418:1 16.66 ± 1.80 17.36 ± 1.68 16.62 ± 0.74 17.65 ± 2.3120:4 n-6 6.64 ± 0.40 6.70 ± 0.33 6.82 ± 0.44 6.56 ± 0.6222:4 n-6 3.74 ± 0.42 3.81 ± 0.36 3.77 ± 0.25 3.82 ± 0.5522:6 n-3 11.16 ± 0.42 11.07 ± 0.70 11.61 ± 0.87 10.20 ± 0.77

The fatty acid content of frontal cortex brain tissue samples from four human subjects was determined by GC analysis of fatty acid methyl esters and the values ± S.E.are in �mol/g wet weight. Repeated measures ANOVA showed no significant differences (P < 0.05) between the levels of fatty acids obtained by the variousprocedures.

F oomE l repr

2

w

ig. 1. Graphs showing the stability of brain fatty acid composition over time at rach graph represents information for a the specified fatty acid and each symbo

.6. Statistical analysis

One-way repeated measures analysis of variance (ANOVA)as used to compare different methods of extraction and saponi-

fiDsy

temperature. Fatty acids were analysed as methyl esters by gas chromatography.esents data from a different brain.

cation of lipids, and preparation of fatty acid methyl esters.ata from a series of brain tissue samples from the four human

ubjects were compared by Bonferroni adjusted pair-wise anal-sis. P values less than 0.05 were considered significant.

138 T. Fraser et al. / Journal of Neuroscience Methods 169 (2008) 135–140

over

3

smatdcasfNdtT(

mohp

scafdcIac

4

ptmt

Fig. 2. Stability of brain tissue fatty acid composition

. Results

Our first comparison was between mechanical homogeni-ation with beads in chloroform-washed 2 ml screw-capicrocentrifuge tubes and manual homogenisation in vitro usingPotter-Elvehjem homogeniser. The antioxidant BHT was used

o protect polyunsaturated fatty acids. There was no significantifference in the levels of the fatty acids extracted by the two pro-edures (Table 1). The results of carrying out the saponificationnd methylation steps at room temperature overnight in smallealed glass vials were then compared with the results obtainedor the two steps carried out for 1 h at 80 ◦C with capping under

2. There was no significant difference in fatty acid contentetermined by these two approaches (Table 1). Finally the impor-ance of capping under N2 at room temperature was investigated.his also had no effect on the measurements of fatty acid content

Table 1).Supported by these data, we next examined the effect of post-

ortem delay. Our analysis was based on the homogenisationf samples in the Bertin Technologies Precellys 24 mechanicalomogeniser (which allows rapid homogenisation of 20 sam-les simultaneously), and the methylation and saponification of

tanp

time at 4 ◦C. Other details are the same as for Fig. 1.

amples overnight at room temperature, avoiding the need forapping under N2. Our analysis was performed on immediatelydjacent samples of frontal cortex from each of the four dif-erent brains. The samples were stored for periods of up to 3ays at room temperature or at 4 ◦C, after which the fatty acidomposition was determined using the method developed above.t was observed that storage for up to 3 days at room temper-ture (Fig. 1) or 4 ◦C (Fig. 2) had no effect on the fatty acidomposition or total fatty acid content (data not shown).

. Discussion

Present findings demonstrate the utility of a high-throughputrocedure for determining the fatty acid composition of brainissue. The procedure involves the use of a multiple-sampleechanical homogeniser and the methylation and saponifica-

ion of samples in small sealed glass vials overnight at room

emperature, without the need for capping under N2. We havelso shown that the fatty acid composition of brain tissue isot affected by conditions that simulate the storage of the bodyost-mortem for up to 72 h.

urosci

baSmiep(aipLi(c

vaoacltt

A

RpC

R

A

B

B

B

B

C

C

C

C

C

F

F

F

F

H

H

K

L

L

L

L

L

M

M

M

M

P

Q

T. Fraser et al. / Journal of Ne

There is contradictory information on the DHA content ofrain tissue in AD. Both Kwee et al. (1991) and Soderberg etl. (1991) recorded a reduction of DHA levels in AD, whereaskinner et al. (1993) found no difference, except in parietal whiteatter, where DHA constituted a higher proportion of the lipids

n AD than controls. Corrigan et al. (1998) found DHA lev-ls to be lower in the phospholipid fraction in AD than controlarahippocampal cortex but not significantly so. Prasad et al.1998) reported that levels of phosphatidylethanolamine (PE)-nd phosphatidylinositol (PI)-derived fatty acids (of which DHAs a major component) were significantly decreased in the hip-ocampus in AD but not elsewhere in the neocortex. Lastly,ukiw et al. (2005) found DHA levels to be significantly reduced

n the CA1 field of the hippocampus and in the “temporal lobe”presumably temporal neocortex) in AD but not in the occipitalortex or thalamus.

The discrepancies between these studies probably reflect theery small number of subjects in each series (<10 per cohort),nd variations in sampling and methodology. Our developmentf a high-throughput, accurate method for analysing the fattycid composition of brain tissue and our demonstration of theonstancy of the findings over conditions simulating relativelyengthy model post-mortem delay should enable us to resolvehe discrepancies by examining a much larger series of brainshan has been practicable using previous methods.

cknowledgments

This work was supported by a grant from Alzheimer’sesearch Trust. Some of the equipment used in the study wasurchased by BRACE (Bristol Research into Alzheimer’s andare of the Elderly).

eferences

ssisi A, Banzi R, Buonocore C, Capasso F, Di Muzio V, Michelacci F, etal. Fish oil and mental health: the role of n-3 long-chain polyunsaturatedfatty acids in cognitive development and neurological disorders. Int ClinPsychopharmacol 2006;21:319–36.

arberger-Gateau P, Letenneur L, Deschamps V, Peres K, Dartigues JF, RenaudS. Fish, meat, and risk of dementia: cohort study. Br Med J 2002;325:932–3.

azan NG. Neuroprotectin D1 (NPD1): a DHA-derived mediator that protectsbrain and retina against cell injury-induced oxidative stress. Brain Pathol2005;15:159–66.

reckenridge WC, Morgan IG, Gombos G. Lipid composition of adultrat-brain synaptosomal plasma-membranes. Biochim Biophys Acta1972;266:695–707.

roadhurst CL, Wang YQ, Crawford MA, Cunnane SC, Parkington JE,Schmidt WF. Brain-specific lipids from marine, lacustrine, or terrestrial foodresources: potential impact on early African Homo sapiens. Comp BiochemPhysiol B: Biochem Mol Biol 2002;131:653–73.

alon F, Lim GP, Morihara T, Yang FS, Ubeda O, Salem N, et al. Dietary n-3polyunsaturated fatty acid depletion activates caspases and decreases NMDAreceptors in the brain of a transgenic mouse model of Alzheimer’s disease.Eur J Neurosci 2005;22:617–26.

alon F, Lim GP, Yang FS, Morihara T, Teter B, Ubeda O, et al. Docosahexaenoic

acid protects from dendritic pathology in an Alzheimer’s disease mousemodel. Neuron 2004;43:633–45.

ole GM, Morihara T, Lim GP, Yang FS, Begum A, Frautschy SA. NSAIDand antioxidant prevention of Alzheimer’s disease: lessons from in vitro andanimal models. Ann NY Acad Sci 2004;1035:68–84.

R

S

ence Methods 169 (2008) 135–140 139

onquer JA, Tierney MC, Zecevic J, Bettger WJ, Fisher RH. Fatty acid anal-ysis of blood plasma of patients with Alzheimer’s disease, other types ofdementia, and cognitive impairment. Lipids 2000;35:1305–12.

orrigan FM, Horrobin DF, Skinner ER, Besson JAO, Cooper MB. Abnormalcontent of n-6 and n-3 long-chain unsaturated fatty acids in the phosphoglyc-erides and cholesterol esters of parahippocampal cortex from Alzheimer’sdisease patients and its relationship to acetyl CoA content. Int J BiochemCell Biol 1998;30:197–207.

arkas T, Kitajka K, Fodor E, Csengeri I, Lahdes E, Yeo YK, et al. Docosa-hexaenoic acid-containing phospholipid molecular species in brains ofvertebrates. Proc Natl Acad Sci USA 2000;97:6362–6.

arooqui AA, Horrocks LA, Farooqui T. Modulation of inflammation in brain:a matter of fat. J Neurochem 2007;101:577–99.

olch J, Lees M, Stanley GHS. A simple method for the isolation and purificationof total lipides from animal tissues. J Biol Chem 1957;226:497–509.

raser TCM, Qi BX, Elhussein S, Chatrattanakunchai S, Stobart AK, LazarusCM. Expression of the isochrysis C18-delta(9) polyunsaturated fatty acidspecific elongase component alters arabidopsis glycerolipid profiles. PlantPhysiol 2004;135:859–66.

ashimoto M, Hossain S, Shimada T, Sugioka K, Yamasaki H, Fujii Y, etal. Docosahexaenoic acid provides protection from impairment of learn-ing ability in Alzheimer’s disease model rats. J Neurochem 2002;81:1084–91.

ossain MS, Hashimoto M, Gamoh S, Masumura S. Antioxidative effects ofdocosahexaenoic acid in the cerebrum versus cerebellum and brainstem ofaged hypercholesterolemic rats. J Neurochem 1999;72:1133–8.

wee IL, Nakada T, Ellis WG. Elevation in relative levels of brain mem-brane unsaturated fatty acids in Alzheimer’s disease: high resolutionproton spectroscopic studies of membrane lipid extracts. Magn Reson Med1991;21:49–54.

arrieu S, Letenneur L, Helmer C, Dartigues JF, Barberger-Gateau P. Nutritionalfactors and risk of incident dementia in the PAQUID longitudinal cohort. JNutr Health Aging 2004;8:150–4.

auritzen L, Hansen HS, Jorgensen MH, Michaelsen KF. The essentiality oflong chain n-3 fatty acids in relation to development and function of thebrain and retina. Prog Lipid Res 2001;40:1–94.

im GP, Calon F, Morihara T, Yang FS, Teter B, Ubeda O, et al. A diet enrichedwith the omega-3 fatty acid docosahexaenoic acid reduces amyloid burdenin an aged Alzheimer mouse model. J Neurosci 2005;25:3032–40.

ukiw WJ, Bazan NG. Survival signalling in Alzheimer’s disease. Biochem SocTrans 2006;34:1277–82.

ukiw WJ, Cui JG, Marcheselli VL, Bodker M, Botkjaer A, Gotlinger K, et al.A role for docosahexaenoic acid-derived neuroprotectin D1 in neural cellsurvival and Alzheimer disease. J Clin Invest 2005;115:2774–83.

ontine KS, Quinn JF, Zhang J, Fessel JP, Roberts LJ, Morrow JD, et al. Iso-prostanes and related products of lipid peroxidation in neurodegenerativediseases. Chem Phys Lipids 2004;128:117–24.

origuchi T, Greiner RS, Salem N. Behavioral deficits associated with dietaryinduction of decreased brain docosahexaenoic acid concentration. J Neu-rochem 2000;75:2563–73.

orris MC, Evans DA, Bienias JL, Tangney CC, Bennett DA, Wilson RS, etal. Consumption of fish and n-3 fatty acids and risk of incident Alzheimerdisease. Arch Neurol 2003;60:940–6.

uskiet FAJ, Fokkema MR, Schaafsma A, Boersma ER, Crawford MA. Isdocosahexaenoic acid (DHA) essential? Lessons from DHA status regula-tion, our ancient diet, epidemiology and randomized controlled trials. J Nutr2004;134:183–6.

rasad MR, Lovell MA, Yatin M, Dhillon H, Markesbery WR. Regional mem-brane phospholipid alterations in Alzheimer’s disease. Neurochem Res1998;23:81–8.

i BX, Fraser T, Mugford S, Dobson G, Sayanova O, Butler J, et al. Production ofvery long chain polyunsaturated omega-3 and omega-6 fatty acids in plants.Nat Biotechnol 2004;22:739–45.

eich EE, Markesbery WR, Roberts LJ, Swift LL, Morrow JD, Montine TJ.Brain regional quantification of F-ring and D-/E-ring isoprostanes and neu-roprostanes in Alzheimer’s disease. Am J Pathol 2001;158:293–7.

alem N, Litman B, Kim HY, Gawrisch K. Mechanisms of action of docosa-hexaenoic acid in the nervous system. Lipids 2001;36:945–59.

1 scienc

S

S

S

T

control study. Br J Nutr 2003;89:483–9.

40 T. Fraser et al. / Journal of Neuro

chwab JM, Chiang N, Arita M, Serhan CN. Resolvin E1 and protectin D1activate inflammation-resolution programmes. Nature 2007;447:869–74.

kinner ER, Watt C, Besson JAO, Best PV. Differences in the fatty acid composi-

tion of the grey and white matter of different regions of the brains of patientswith Alzheimer’s disease and control subjects. Brain 1993;116:717–25.

oderberg M, Edlund C, Kristensson K, Dallner G. Fatty acid composi-tion of brain phospholipids in aging and in Alzheimer’s disease. Lipids1991;26:421–5.

V

Y

e Methods 169 (2008) 135–140

ully AM, Roche HM, Doyle R, Fallon C, Bruce I, Lawlor B, et al. Low serumcholesteryl ester-docosahexaenoic acid levels in Alzheimer’s disease: a case-

alentine RC, Valentine DL. Omega-3 fatty acids in cellular membranes: aunified concept. Prog Lipid Res 2004;43:383–402.

oung G, Conquer J. Omega-3 fatty acids and neuropsychiatric disorders.Reprod Nutr Dev 2005;45:1–28.