Archives of Medical Research 38 (2007) 489e494

ORIGINAL ARTICLE

Oral Administration of Avocado Soybean Unsaponifiables (ASU) ReducesIschemic Damage in the Rat Hippocampus

Mehmet Yaman,a Olcay Eser,b Murat Cosar,b Orhan Bas,c Onder Sahin,d Hakan Mollaoglu,e

Huseyin Fidan,f and Ahmet Songurc

aDepartment of Neurology, bDepartment of Neurosurgery, cDepartment of Anatomy, dDepartment of Pathology,eDepartment of Physiology, and fDepartment of Anesthesiology, School of Medicine, Kocatepe University, Afyonkarahisar, Turkey

Received for publication August 19, 2006; accepted January 22, 2007 (ARCMED-D-06-00358).

Background. The beneficial effects of avocado/soybean unsaponifiables (ASU) areknown as an antiarthritic agent. This experimental study presents the effects of ASUon oxidant/antioxidant systems and the number of apoptotic neurons of hippocampalformation after ischemia and reperfusion.

Methods. Eighteen rats were divided into three equal groups: group I rats were used ascontrols; group II rats were fed with standard diet and group III rats were fed with stan-dard diet plus ASU pills for 10 days. One day after electrocauterization of bilateral ver-tebral arteries for groups II and III, bilateral common carotid arteries were occluded for30 min and then reperfused for 30 min. After these procedures, rats of all groups weresacrificed. The levels of malondialdehyde (MDA) and nitric oxide (NO) and activitiesof superoxide dismutase (SOD) and catalase (CAT) were measured in the left hippocam-pus. The number of apoptotic neurons was counted by Tunel method in histologicalsamples of right hippocampus.

Results. MDA and NO levels increased in group II compared with group I rats ( p 50.002, p 5 0.015). In group III, MDA and NO levels decreased as compared to groupII ( p 5 0.041, p 5 0.002). SOD and CAT activities increased in group III as comparedto group II rats ( p 5 0.002, p 5 0.002). The number of apoptotic neurons was lower ingroup III as compared to group II rats.

Conclusions. The present findings suggest that ASU could decrease oxidative stress andapoptotic changes in ischemic rat hippocampus. Dietary supplementation of ASU may bebeneficial to prevent or ameliorate ischemic cerebral vascular disease. � 2007 IMSS.Published by Elsevier Inc.

Key Words: ASU, Cerebral/Ischemia/Reperfusion, Hippocampus.

Introduction

Cerebral ischemia or stroke is one of the leading causes ofdeath and long-term disability in aged populations, often re-sulting in irreversible brain damage and subsequent loss ofneuronal function. Cerebral ischemia results in low oxygenand glucose supply and a decrease in adenosine triphos-phate (ATP) formation. The brain is very susceptible toenergy-depriving injuries and is particularly sensitive tooxygen radical-mediated injury due to its characteristics

Address reprint requests to: Mehmet Yaman, Ornekevler MH, Gelincik

sk. No. 28/10 03200, Afyonkarahisar, Turkey; E-mail: yaman@aku.edu.tr

0188-4409/07 $esee front matter. Copyright � 2007 IMSS. Published by Elsdoi: 10.1016/j.arcmed.2007.01.008

of low fuel reserves, high aerobic metabolism, low concen-trations of oxygen radical scavenging enzymes, and highlevels of polyunsaturated fatty acids that constitute the neu-ronal cell membrane (1). It is known that ischemia en-hances the formation of reactive oxygen species (ROS) inbrain tissue. ROS, such as superoxide anions, hydroxyl freeradicals, hydrogen peroxide and nitric oxide, are producedas a consequence of metabolic reactions and central ner-vous system activity (2). ROS are directly involved in oxi-dative damage of cellular macromolecules such as nucleicacids, proteins and lipids in ischemic tissues, which canlead to cell death (3). ROS are scavenged by superoxidedismutase (SOD), glutathione peroxidase and catalase.

evier Inc.

490 Yaman et al. / Archives of Medical Research 38 (2007) 489e494

Another small antioxidant molecule, glutathione (GSH), isinvolved in the detoxification of free radicals (4). Indeed,overproduction of ROS and consequent oxidative stresshas a critical role in the development of brain injury (5).The balance between cell death and proliferation/cell sur-vival determines the characteristics of neuronal cell nuclearvolumes in various brain regions.

Avocado/soybean unsaponifiables (ASU) is a new antiar-thritic agent derived from unsaponifiable residues of avo-cado and soya oils, mixed at a ratio of 1:2, respectively.ASU has recently been shown in various in vitro systemsto partially reverse the effects of interleukin-1 (IL-1) oncultured chondrocytes, stimulating collagen synthesis andinhibiting the production of matrix metalloproteases, IL-6,IL-8 and prostaglandin E2 (PGE2) (6). A similar effecthas also been demonstrated in rheumatoid synoviocytes(7). The active ingredients of the plant extract are unknown,but synergism between the avocado and soya componentsand their relative ratios appears to be important (6,8).ASU is manufactured in France under the name Piascle-dine, and to our knowledge, distributed only in France.The beneficial effects of ASU on osteoarthritis (OA) man-agement was reported in previous studies (8); however, thisis the first study about its effects on cerebral ischemia in-jury. To this end, in this study we investigated the neuro-protective effects of oral administration of ASU in thehippocampus after ischemia/reperfusion (I/R) injury.

Materials and Methods

Animals and Care

Experiments were performed in accordance with the ‘‘Ani-mal Welfare Act’’ and the Guide for the Care and Use ofLaboratory Animals, prepared by the Kocatepe University,Animal Ethical Committee. Female Sprague Dawley (Har-lan, Indianapolis, IN) rats (aged 8e12 weeks) weighing300 � 20 g (mean � SD) were used in the experiments.Rats that have the same menstrual cycles detected with vag-inal smear investigation were included in the study. The ratswere placed in a temperature- (21 � 2�C) and humidity-(60 � 5%) controlled room in which a 12 h:12 h light:darkcycle was maintained for 1 week before the experiment. Astandard diet (without any unsaponifiable residues of avo-cado and soya oil, and soybean, and avocado product) (AfyonYem Sanayi Ltd., Afyonkarahisar, Turkey) and tap waterwere provided ad libitum.

Rats were randomly divided into three experimentalgroups as follows: group I was control and nonischemicrats (n 5 6); group II was ischemic rats fed on a standarddiet (n 5 6); group III was ischemic rats fed on the standarddiet plus ASU dissolved in 1 mL flower oil (Piascledine300, Expanscience Lab, France, 0.6 g/kg/day, by gavage)for 10 days (n 5 6). In the control and ischemic groups,

sunflower oil was given in the same manner for 10 days.The last dose of ASU was administrated 12 h before theI/R procedure.

Experimental Groups and Surgical Procedures

After 10 days, rats were anesthetized with IP ketamine-xylazine (50 mg/kg�6 mg/kg) combination. For group Irats, the neck incision was left open for 30 min, but thecommon carotid arteries (CCA) and vertebral arteries werenot occluded. This group of animals was used to determinethe effects of anesthesia and the operation on the results.

Group II and III rats were prepared for transient globalcerebral ischemia using a four-vessel occlusion model (9).On day 1, both vertebral arteries were electrocauterized atthe level of the first cervical vertebra within the alar foram-ina. After the closure of incision, the rats were wakened.

On the next day (24 h later), the rats were again anesthe-tized. The bilateral femoral vein and artery were canalizedwith a cannula (#24) (Datex/Ohmeda S/5, Helsinki, Fin-land) by inguinal incision. Arterial pressure and arterialpulses of the rats were monitored and measured with a can-nula. In both groups II and III, the I/R model was performedvia occlusion of the bilateral CCA. For this purpose, bilat-eral CCA were exposed with a transverse ventral neck inci-sion and bilateral CCA were occluded with aneurysm clips(Sugita, temporary aneurysm clip, Mizuho, Japan) for 30min, followed by 30 min of reperfusion. A BiopacMP150 device was used as a monitor. During the ischemicperiod, the mean arterial pulse was 385/min and body tem-perature was maintained with a heating pad at 36.5e37.5�Cas monitored via a rectal probe.

At the end of the procedure, the rats were sacrificed andbrains removed immediately. The left hippocampus wasdissected and washed with ice-cold isotonic NaCl to re-move residual blood. Tissues were stored at �70�C untilbiochemical analysis. The right hemispheres were blockedfor histopathological analysis.

Biochemical Analysis

Determination of protein content. The protein content oflung and brain homogenates was measured by the methodof Lowry et al. (10) with bovine serum albumin as the stan-dard. A Shimadzu spectrophotometer (Shimadzu Corp.,Kyoto, Japan) was used for all assays.

Measurement of MDA levels in hippocampal tissue. Deter-mination of MDA levels was based on the coupling ofMDA with thiobarbituric acid at þ95�C (11). Results wereexpressed as nanomole per gram wet tissue (nmol/g wettissue).

Measurement of nitrite production in hippocampal tissue.The method for measuring plasma nitrite and nitrate levelswas based on the Griess reaction (12). Samples were

491ASU and Cerebral Ischemia

initially deproteinized with Somogyi reagent. Total nitrite(nitrite þ nitrate) was measured by spectrophotometry at545 nm after conversion of nitrate to nitrite by copperizedcadmium granules. Results were expressed as micromoleper gram tissue protein (mmol/g protein).

Measurement of SOD activity in hippocampal tissue. TissueSOD activity was determined using the nitroblue tetrazo-lium (NBT) method described by Sun et al. (13) and mod-ified by Durak et al. (14). In this method, NBT is reduced toblue formazan by superoxide (O2

�), which has a strong ab-sorbance at 560 nm. One unit (U) of SOD is defined as theamount of protein that inhibits the rate of NBT reduction by50%. The calculated SOD activity is expressed as U/mgprotein.

Measurement of catalase activity in hippocampal tissue.Catalase (CAT, EC 1.11.1.6) activity was determined ac-cording to Aebi’s method (15). The principle of the assayis based on determination of the rate constant k (s�1) ofthe H2O2 decomposition rate at 240 nm. Results wereexpressed as k (rate constant) per gram of protein (k g�1

protein).

Histopathologic Procedures

TUNEL staining. After decapitation, the right hemispheresof the brains were fixed in 4% paraformaldehyde inphosphate-buffered saline (PBS), dehydrated through etha-nol and xylene, and embedded in paraffin. Five-mm-thicksagittal sections through the right hemisphere were obtainedand mounted on poly-L-lysine-coated glass slides. The Apop-Tag peroxidase kit (Oncor, Gaithersburg, MD) was appliedon paraffin sections according to the manufacturer’s instruc-tions. Briefly, residues of digoxigenin nucleotide were cata-lytically added by TdT to the 30-OH ends of double- orsingle-stranded DNA. The labeled product was visualized us-ing diaminobenzidine (DAB), which yielded brown granulesmainly localized to apoptotic cells. Following this, the sec-tions were counterstained with 1% methyl green. Omissionof TdT or digoxigenin-nucleotide gave completely negativeresults (data not shown).

Light microscopy analysis. Comparisons of levels ofneuronal cell death were determined from the number ofmorphologically intact cells and the number of deep-brown-stained, TUNEL-positive, damaged cells in thehippocampal CA1, CA2, CA3 and dentate gyrus (Dg) re-gions. Preparations were imaged using a brightfield micro-scope (Nikon Microscope Eclipse E600W, Tokyo, Japan)and were photographed with a microscope digital camera(Nikon DP70).

The number of cells in each group was counted sepa-rately in four hippocampal areas per section (two areas eachper portion). An area was defined 10 m as 255 mm (W) �

165 mm (H), with the neuronal cell layer of the hippocam-pal CA1, CA2, CA3 and Dg regions calculated by an imageanalysis system. The investigator who performed thesemeasurements was unaware of the experiment. The systemused was composed of a PC, hardware and software cam-era. Images were processed by an IBM-compatible personalcomputer, high-resolution video monitor and image analy-sis software (BS200Docu Version 2.0; BAB Imaging Sys-tems, Ankara, Turkey) and an optical microscope. Themethod required preliminary software procedures of spatialcalibration (micron scale) and color segmentation for quan-titative color analysis set up parallel to the major axis. Thenumber of cells was calculated as an average per mouse. Anexample of a cell-counting area and typical images of dam-aged and undamaged neuronal cells are shown in Figure 1.

Statistical Analysis

All biochemical and histopathological data were presentedas mean � standard deviation (SD). A computer program(SPSS 9.0; SPSS Inc., Chicago, IL) was used for statisticalanalysis. Data were considered to be non-parametric; there-fore, they were analyzed using a KruskaleWallis H test.Differences between two groups were determined witha ManneWhitney U test; p !0.05 was considered statisti-cally significant.

Results

Biochemical Analysis

MDA levels. As shown in Table 1, MDA levels were 45.38� 10.26 (mean � SD) in the sham-operated group (groupI), 94.96 � 6.93 in the I/R þ untreated group (group II),and 73.31 � 18.06 in the I/R þ ASU group (group III). Ac-cording to these values, MDA levels increased significantlyfrom group I to group II ( p 5 0.002). The MDA levelincrease was prevented significantly with ASU treatmentwhen compared between group II and group III ( p 5

0.041).

Catalase activity. As shown in Table 1, catalase activitieswere 0.334 � 0.050 (mean � SD) in the sham-operatedgroup (group I), 0.114 � 0.017 in the I/R þ untreated group(group II), and 0.439� 0.086 in the I/RþASU group (groupIII). Catalase activity decreased significantly in group II ascompared to group I ( p 5 0.002). Catalase activity increasedsignificantly in the ASU-treated group (group III) ascompared to the untreated group (group II) ( p 5 0.002).

SOD activity. As seen in Table 1, SOD activities were0.827 � 0.136 (mean � SD) in the sham-operated group(group I), 0.648� 0.097 in the I/Rþ untreated group (groupII), and 3.98 � 0.609 in the I/R þ ASU group (group III).

492 Yaman et al. / Archives of Medical Research 38 (2007) 489e494

Figure 1. TUNEL staining of sagittal sections of rat hippocampus of control. (A) (CA1, CA1, CA2, CA3 and Dg); ischemic-untreated rat hippocampus. (B)

(CA1, CA1, CA2, CA3 and Dg); and ischemic plus ASU, C (CA1, CA1, CA2, CA3 and Dg), in which TUNEL-positive cells had markedly decreased. Brown

nuclei represent TUNEL-labeled cells. Nuclei were counterstained with methyl green. Bar 5 200 mm.

Most notably, in the ASU-treated group (group III), SODactivities increased significantly as compared to theuntreated group (group II) ( p 5 0.002). Differences inSOD activities among the three groups were also significant( p !0.05).

NO levels. As shown in Table 1, NO levels were 5.92 �1.99 (mean � SD) in the sham-operated group (group I),8.91 � 1.56 in the I/R þ untreated group (group II), and3.05 � 0.69 in the I/R þ ASU group (group III). Although

NO levels increased significantly in the untreated group(group II) as compared to the sham group (group I) ( p 5

0.015), NO levels decreased significantly with ASU treat-ment (group III) as compared to group II ( p 5 0.002).

Histopathology. Under the light microscope, almost allneurons in the CA1, CA2, CA3 and in Dg subfields ofthe hippocampus of group I were normal in appearance.The number of surviving neurons was markedly decreasedin group II, particularly in the CA1 and CA2 subfields. The

493ASU and Cerebral Ischemia

Table 1. Result of ischemia/ASU biochemistry*

Groups MDA (nmol/g wet tissue) NO (mmol/g protein) SOD (U/mg protein) CAT (k/g protein)

I (sham-operated, n 5 6) 45.38 � 10.26 5.92 � 1.99 0.827 � 0.136 0.334 � 0.050

II (ischemia þ untreated rats, n5 6) 94.96 � 6.93 8.91 � 1.56 0.648 � 0.097 0.114 � 0.017

III (ischemia þ ASU, n 5 6) 73.31 � 18.06 3.05 � 0.69 3.980 � 0.609 0.439 � 0.086

p values

IeII 0.002 0.015 0.026 0.002

IeIII 0.002 0.026 0.002 NS

IIeIII 0.041 0.002 0.002 0.002

NS: Not significant ( p !0.05).

*Levels of malonedialdehyde (MDA) and tissue nitric oxide (NO), and activities of superoxide dismutase (SOD) and catalase (CAT) in control, ischemic

and ischemic plus ASU groups in rats.

number of apoptotic cells in all subfields of the hippocam-pus decreased in group III. These results suggest that thenumber of apoptotic cells increases in the hippocampusafter I/R injury, but oral ASU administration decreasesapoptosis in these regions (Table 2; Figure 1).

Discussion

Stroke causes brain injury in millions of people worldwideeach year (16). Despite the enormity of the problem, thereis no approved therapy that can reduce this neurologicaldisability. Recently, both animal and human studies haveprovided evidence that oxidative damage to membranelipids and proteins increases during I/R (3). Explorationof oxidative stress is important for evolving neuroprotectivestrategies to enhance neuronal survival after cerebral ische-mia. Thus, the aim of experimental studies must focuson the development of molecules that decrease oxidativedamage.

ROS can cause cellular damage by oxidizing membranelipids, essential cellular proteins and DNA (4). MDA is oneof the most sensitive indicators of lipid peroxidation (17).As long as fatty acids, O2 and metal catalysts (Feþ2,Cuþ) are present, lipid peroxidation leads to the formationof new free radicals. Therefore, the period of reperfusion ishighly suitable for lipid peroxidation (18). The impairment

Table 2. Number of apoptotic cells in subfields of hippocampus (CA1,

CA2, CA3, Dg) of sham, ischemic and ischemic þ ASU groups

Groups CA1 CA2 CA3 Dg

I (sham-operated, n 5 6) 0.5 � 0.8 0.3 � 0.5 0.3 � 0.5 0.7 � 0.8

II (ischemia þ untreated

rats, n 5 6)

12.3 � 3.3 12.3 � 2.6 9.2 � 2.0 27.3 � 3.3

III (ischemia þ ASU,

n 5 6)

4.8 � 1.3 6.7 � 1.0 5.5 � 1.0 14.3 � 1.7

p values

IeII 0.003 0.003 0.003 0.004

IeIII 0.003 0.003 0.003 0.004

IIeIII 0.004 0.004 0.004 0.004

in membrane permeability due to lipid peroxidation cancause a decrease in membrane-bound Naþ-KþATPase en-zyme activity. As a result, the Kþ and Mg2þ concentrationsvitally important for protein synthesis are altered, and pro-tein synthesis is inhibited. Increased lipid peroxidation canalso result in the release of proteolytic lysosomal enzymesand mitochondrial matrix enzymes into the cytoplasm,which gives rise to intracellular proteolysis and cellular de-struction (18). Under these conditions, the defensive sys-tem, which includes such antioxidant enzymes as SOD,has a crucial role in the resistance of the neuronal cells toROS-induced cell death (4). SOD specifically detoxifiesO2� to H2O2, which is then scavenged by peroxisomal cat-

alase (4). In short, during I/R, H2O2 cannot be readily scav-enged because of low activities of SOD, CAT, andglutathione peroxidase (19).

In our study, tissue MDA levels increased significantlyafter I/R (group II) as compared to the sham-operated group(group I) ( p 5 0.002). SOD ( p 5 0.026) and CAT ( p 5

0.002) activities decreased significantly after I/R (groupII) as compared to the sham-operated group. In group III,which was treated with ASU before I/R, MDA levelsdecreased significantly as compared to the control I/Rgroup ( p 5 0.041) (Group II). In addition to these data,the significant increase in SOD ( p 5 0.002) and CAT activ-ities ( p 5 0.002) in group III as compared to group IIpointed to possible neuroprotective effects of ASU in thehippocampal region.

It is known that tissue NO synthesis increases in somepathological conditions. A high concentration of NO couldinduce apoptotic cell death. Many pathways, such as inter-actions with excitatory amino acid receptors, the depletionof cellular NADþ (20) and the activation of caspase (21),are involved in the cascade of events leading to NO-inducedapoptosis (20,21). We showed that oral administration ofASU for 10 days before I/R injury decreases tissue NOproduction and apoptosis in the hippocampus.

The published animal study found that oral ASU signif-icantly prevented the occurrence of cartilage lesions ina rabbit post-contusive model (22). While its symptomaticefficacy has been demonstrated recently in double-blind,

494 Yaman et al. / Archives of Medical Research 38 (2007) 489e494

placebo-controlled prospective clinical trials in patientswith OA, (8,23) including evidence of reduced joint spacenarrowing (24), its ability to also protect the articular carti-lage of osteoarthritic joints in vivo remains to be estab-lished. The result of our study showed that the effects ofASU on hippocampal tissue of rats was similar to the re-sults on cartilage and bone tissues. Additionally, the roleof avocado derivatives as anticonvulsant agents has recentlybeen published, probably exerted by means of GABAcurrents potentiation (25).

In conclusion, this study showed that oral administrationof ASU has a neuroprotective role during brain ischemiafollowed by reperfusion in rats. The neuroprotection shownby ASU may be attributed to inhibition of lipid peroxida-tion, an increase in endogenous antioxidant defense en-zymes and a reduction in tissue NO formation. Thisimplies that consumption of ASU may have beneficial ef-fects for those who are at risk for cerebrovascular disease.Further work is required to determine if these findingscan be extended to prophylactic use of ASU for cerebro-vascular disease.

AcknowledgmentsThis study was supported by the Kocatepe University ScientificResearch Committee (Project #051.TIP.14).

References1. Tian J, Fu F, Geng M, Jiang Y, Yang J, Jiang W, et al. Neuroprotective

effect of 20(S)-ginsenoside Rg3 on cerebral ischemia in rats. Neurosci

Lett 2005;374:92e97.

2. Lanselot E, Callebert J, Revaud ML, Boulu RG, Plotkine M. Detection

of hydroxyl radicals in rat striatum during transient focal cerebral is-

chemia: Possible implication in tissue damage. Neurosci Lett 1995;

197:85e88.

3. Chan PH. Reactive oxygen radicals in signalingand damage in the

ischemic brain. J Cereb Blood Flow Metab 2001;21:2e14.

4. Ginsberg MD. Neuroprotection in brain ischemia: an update (parts 1

and 2). Neuroscientist 1995;1:95.

5. Cao W, Carney JM, Duchon A, Floyd RA, Chevion M. Oxygen free

radicals involvement in ischemia and reperfusion injury to brain.

Neurosci Lett 1998;88:233e238.

6. Henrotin YE, Labasse AH, Jaspar JM, de Groote DD, Zheng SX,

Guillou GB, et al. Effects of three avocado/soybean unsaponifiable

mixtures on metalloproteinases, cytokines and prostaglandin E2

production by human articular chondrocytes. Clin Rheumatol 1998;

17:31e39.

7. Mauviel A, Loyau G, Pujol JP. Effet des insaponifiables d’avocat/soja

(Piascle dine) sur l’activite collage nolytique de cultures de synovio-

cytes rhumatoides humains et de chondrocytes articulaires de lapin

traite’s par l’interleukine-1. Rev Rhum 1991;58:241e245.

8. Maheu E, Mazieres B, Valat JP, Loyau G, le Loet X, Bourgeois P, et al.

Symptomatic efficacy of avocado/soybean unsaponifiables in the treat-

ment of osteoarthritis of the knee and hip. A prospective, randomized,

double-blind, placebo-controlled, multicenter clinical trial with

six-month treatment period and a two-month followup demonstrating

a persistent effect. Arthritis Rheum 1998;41:81e91.

9. Onem G, Aral E, Enli Y, Oguz EO, Coskun E, Aybek H, et al. Neuro-

protective effects of L-carnitine and vitamin E alone or in combination

against ischemia-reperfusion injury in rats. J Surg Res 2006;131:

124e130.

10. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein measure-

ment with the Folin phenol reagent. J Biol Chem 1951;193:265e275.

11. Wasowicz W, Neve J, Peretz A. Optimized steps in fluorometric deter-

mination of thiobarbituric acid-reactive substances in serum: impor-

tance of extraction pH and influence of sample preservation and

storage. Clin Chem 1993;39:2522e2526.

12. Cortas NK, Wakid NW. Determination of inorganic nitrate in serum

and urine by a kinetic cadmium-reduction method. Clin Chem 1990;

36:1440e1443.

13. Sun Y, Oberley LW, Ying L. A simple method for clinical assay of

superoxide dismutase. Clin Chem 1988;34:497e500.

14. Durak I, Yurtarslani Z, Canbolat O, Akyol O. A methodological ap-

proach to superoxide dismutase (SOD) activity assay based on inhibi-

tion of nitroblue tetrazolium (NBT) reduction. Clin Chim Acta 1993;

214:103e104.

15. Aebi H. Catalase. In: Bergmeyer HU, ed. Methods of Enzymatic Anal-

ysis. New York: Academic Press; 1974. pp. 673e677.

16. Read JS, Hirano T, Davis SM, Donnan GA. Limiting neurological

damage after stroke. Drugs Aging 1999;14:11e39.

17. Gutteridge JM. Lipid peroxidation and antioxidants as biomarkers of

tissue damage. Clin Chem 1995;41:1819e1828.

18. White BC, Grossman LI, Krause GS. Brain injury by global ischemia

and reperfusion: a theoretical perspective on membrane damage and

repair. Neurology 1993;43:1656e1665.

19. Sinet PM, Heikkila RE, Cohen G. Hydrogen peroxide production by

rat brain in vivo. J Neurochem 1980;34:1421e1428.

20. Mebmer UK, Brune B. NO in apoptotic versus necrotic RAW 264.7

macrophage cell death: the role of NO-donor exposure, NAD content,

and p53 accumulation. Arch Biochem Biophys 1996;327:1e10.

21. Bosca L, Hortelano S. Mechanisms of NO-dependent apoptosis:

involvement of mitochondrial mediators. Cell Signal 1999;11:

239e244.

22. Mazieres B, Tempesta C, Thiechard M, Vaguier G. Pathologic and

biochemical effects of a lipidic avocado and soy extract (LASE) on

an experimental post-contusive model of OA. Osteoarthritis Cart

1993;1:46e47.

23. Blotman F, Maheu E, Wulwik A, Caspard H, Lopez A. Efficacy and

safety of avocado/soybean unsaponifiables in the treatment of symp-

tomatic osteoarthritis of the knee and hip. A prospective, multicenter,

threemonth, randomized, double-blind, placebo-controlled trial. Rev

Rhum 1997;64:825e834.

24. Lequesne M, Maheu E, Cadet C, Caspard H, Dreiser RL. Effect of

avocado/soya unsaponifiables (ASU) on joint space loss in hip osteo-

arthritis (HOA) over 2 years. A placebo controlled trial. Arthritis

Rheum 1996;39(Suppl 9):S227.

25. Ojewole JAO, Amabeoku GJ. Anticonvulsant effect of Persea ameri-

cana Mill (Lauraceae) (avocado) leaf aqueous extract in mice. Phyt-

other Res 2006;20:696e700.