Available online at www.sciencedirect.com

ScienceDirect

istry xx (2014) xxx–xxx

Journal of Nutritional Biochem

Maternal hydroxytyrosol administration improves neurogenesis and cognitivefunction in prenatally stressed offspring

Adi Zhenga, Hao Lia, Ke Caob, Jie Xua, Xuan Zoub, Yuan Lia, Cong Chena, Jiankang Liua, Zhihui Fengb,⁎

aCenter for Mitochondrial Biology and Medicine, The Key Laboratory of Biomedical Information Engineering of Ministry of Education, School of Life Science and Technology, Xi'an JiaotongUniversity, Xi'an, China

bCenter for Mitochondrial Biology and Medicine, Frontier Institute of Science and Technology and School of Life Science and Technology, Xi'an Jiaotong University, Xi'an, China

Received 11 April 2014; received in revised form 1 August 2014; accepted 7 October 2014

Abstract

Prenatal stress is known to induce emotional and cognitive dysfunction in the offspring of both humans and experimental animals. Hydroxytyrosol (HT), amajor polyphenol in olive oil with reported ability modulating oxidative stress and mitochondrial function, was performed to investigate its preventive effect onprenatal stress-induced behavioral and molecular alterations in offspring. Rats were exposed to restraint stress on days 14–20 of pregnancy. HT was given atdoses of 10 and 50 mg/kg/day. The spontaneous alternation performance and Morris water maze confirmed the impaired learning capacity and memoryperformance induced by prenatal stress in both male and female offspring, and these effects were markedly restored in the HT supplement groups. Throughtissue analysis of the hippocampi of male offspring, we found that the stress-induced downregulation of neural proteins, including BDNF, GAP43, synaptophysin,NMDAR1, NMDANR2A and NMDANR2B, was prevented by HT. Prenatal stress-induced low expression of glucocorticoid receptor was also increased by HT,although basal fetal serum corticosterone levels were not different among the four groups. Oxidative stress and mitochondrial dysfunction in prenatally stressedrats were confirmed with changes in protein oxidation, SOD activity, the expression of mitochondrial complexes and mitochondrial DNA copy number.Meanwhile, HT significantly increased transcription factors FOXO1 and FOXO3, as well as phase II enzyme-related proteins, including Nrf2 and HO-1, which maycontribute to the decreased oxidative stress and increased mitochondrial function shown with HT supplementation. Taken together, these findings suggest thatHT is an efficient maternal nutrient protecting neurogenesis and cognitive function in prenatally stressed offspring.© 2014 Elsevier Inc. All rights reserved.

Keywords: Prenatal stress; Cognitive dysfunction; Hydroxytyrosol; Oxidative stress; Mitochondrial function

1. Introduction

Stress can occur throughout the entire lifespan. Acute stress maybe initially adaptive to a new allostasis, but excessive, repeated orchronic stress, especially during critical phases of development, mayhave detrimental long-term effects on body functions. Early lifeexperiences, including those in utero, can have long-term conse-quences for the unborn child. Recent findings in human studies showthat prenatal stress leads to maladaptive consequences for theoffspring, including negative birth outcomes [1], altered physiologicalstress responses, behavior problems [2,3] and impaired cognitive andmotor development [4]. Prenatal stress may also increase the risk forseveral adult onset chronic diseases, including cardiovascular disease,type 2 diabetes, obesity and hypertension [5,6]. Because the brainplays a key role in the processing of stressors and in the regulation of

⁎ Corresponding author at: Center for Mitochondrial Biology andMedicine, FIST, Xi'an Jiaotong University, 28 W, Xian-ning Road, Xi'an710049, China. Tel./fax: +86 29 82665849.

E-mail address: zhfeng@mail.xjtu.edu.cn (Z. Feng).

http://dx.doi.org/10.1016/j.jnutbio.2014.10.0060955-2863/© 2014 Elsevier Inc. All rights reserved.

the ensuing behavioral and physiological stress responses [7], studiesin animal experiments also suggest that prenatal stress impairs braindevelopment, strengthens behavioral abnormalities and provokesneurological diseases such as depression and schizophrenia inoffspring [8,9].

Maternal nutritional factors during pregnancy have been linked tofetal brain development and subsequent offspring behavior. Maternalsupplementation with docosahexaenoic acid or wolfberry showedprotective effects on learning and memory, oxidative status andmitochondrial function in prenatally stressed rats [10,11]. Recentreports have indicated that the protective role of aMediterranean dietduring pregnancy for the health of the mother and child is mainly dueto the antioxidants supplied by components of this diet, such as oliveoil [12,13]. Hydroxytyrosol (HT), one of the most effective antioxi-dants, is a natural, well-known phenolic compound from virgin oliveoil. It has been reported that HT has multiple biological functions,including anticancer [14], anti-inflammation [15] and neuroprotec-tive effects [16,17]. Our previous study also demonstrated that HTcould activate phase II enzymes and mitochondrial biogenesis topromote cell survival [18–20]. Despite the limited reports of HTprotection on neurons, the in vivo effects of HT on cognitive functionin prenatally stressed rats remain unknown.

Fig. 1. Effects of prenatal stress and HT supplementation on offspring cognitive function in the T-maze test. Alternation behavior percentages in male rats (A) and female rats (B). Totalnumber of goal arm entries in male rats (C) and female rats (D). Values are mean values±S.E.M. n≥10, ⁎Pb.05, ⁎⁎Pb.01.

2 A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

Prenatal restraint stress (PRS) in rats is one of the most commonmodels of early stress known to induce long-lasting neurobiologicaland behavioral alterations, including impaired feedback mechanismsof the hypothalamo-pituitary-adrenal (HPA) axis, altered neuroplas-ticity and cognitive dysfunction [21]. Further, the hippocampus issensitive to stress, and it is an important component of the stressresponse, which is involved in the regulation of learning andmemory.Therefore, in the current study, we aimed to investigate the potentialbenefit of HT on brain function with a PRS model. We particularlyfocused on the protective effects of maternal HT administration oncognitive function and biochemical function in the hippocampus,including HPA, oxidative stress andmitochondrial dysfunction, whichcan contribute to the pathophysiology of both stress and neurode-generative disorders. Here, we first determined the protectionfunction of HT on learning and memory to prenatally stressed ratsand explored its potential mechanism.

2. Materials and methods

2.1. Reagents and antibodies

The anti-actin antibody was obtained from Sigma (St. Louis, MO, USA). Antic-omplex I, II, III, IV, and V antibodies were obtained from Invitrogen (Carlsbad, CA, USA).Anti-brain-derived neurotrophic factor (BDNF), anti-Forkhead box protein (FOX) O1,anti-FOXO3, anti-HO-1, anti-Nrf2and anti-SOD2 antibodies were from Santa Cruz(Santa Cruz, CA, USA). The Reverse Transcription System kit was purchased fromPromega (Mannheim, Germany). SYBR green was purchased from Takara (Otsu,Japan). Polymerase chain reaction (PCR) primers were synthesized by Baiaoke Biotech(Beijing, China). TRIzol and other reagents were purchased from Invitrogen (Carlsbad,CA, USA). HT was purchased from APP-Chem Bio (Xi'an, China).

2.2. Animals

Specific pathogen-free (SPF) Sprague–Dawley rats were purchased from acommercial breeder (SLAC, Shanghai, China). The rats were housed in a temperature(23–26°C)- and humidity (60%)-controlled animal room and maintained on a 12-hlight/12-h dark cycle (light from 08:00 a.m. to 08:00 p.m.) with food and waterprovided during the experiments. Female rats weighing from 230–250 g and male ratsweighing 280–350 g were used. At the beginning, female rats were randomly assignedto the following experimental groups: control, stress, HT treatment groups (10 and50 mg/kg/day). HT was administered through gavage 2 weeks before mating, and thedoses were chosen based on our previous study [22]. The day on which a vaginal smearwas determined to be sperm-positive was set as embryonic day 0. Each pregnant ratwas then housed separately.

2.3. Stress procedure

The stress protocol was following our previous study [11]. The pregnant rats in thestress andHT treatment groupswereexposed to restraint stressduringembryonic days14to 20. The restraint was performed using a transparent plastic tube of 6.8 cm in diameterfor 2 h during the time periods 08:00 a.m. to 11:00 a.m., 11:00 a.m. to 2:00 p.m., and04:00 p.m. to 07:00 p.m.. The pregnant rats in the control groupwere left undisturbed. Alloffspring were weaned on day 21 after birth, and female and male pups were housedseparately until testing at 1 month of age. All procedures were carried out in accordancewith theUnitedStatesPublicHealth ServicesGuide forCareandUseof LaboratoryAnimalsand were approved by the Institutional Animal Care and Use Committee at School of LifeScience and Technology, Xi'an Jiaotong University. All efforts were made to minimize thenumber of animals used and their suffering.

2.4. Spontaneous alternation performance in the T-maze

Spontaneous alternation is the phenomenon that rats and mice in a T maze have anatural tendency to alternate choice of goal arm (left or right arm) during exploration.The size of the maze was referring the parameters for rats in a previous publication[23]. The rats were confined to the start area for 30 s before testing; then, the sliding

Fig. 2. Effects of prenatal stress and HT supplementation on offspring cognitive function in the Morris water maze. Escape latencies in the MWM in male offspring (A) and femaleoffspring (B). Probe test crossings in male offspring (C) and female offspring (D). All groups of animals were able to learn the task through consecutive trials. The latency per testingsession represents the average of three trials for all animals in each group. Values are mean values±S.E.M. Females, n≥12 for each group; males, n≥15 for each group. Statisticalsignificance of differences was determined by two-way ANOVA (general linear model) repeated measures followed by LSD post hoc analysis, ⁎Pb.05, ⁎⁎Pb.01.

3A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

door was removed, and the rats were allowed to freely explore the rest of the maze foran 8-min trial [24,25]. Numbers of visits to each goal arm and spontaneous alternationperformance were recorded by the observer. The apparatus was cleaned after each ratto remove any residual odors and debris.

2.5. Morris water maze

The Morris water maze consisted of a black circular pool (180 cm diameter, 50 cmhigh), filled with water (21–23°C), mounted with four different visual cues at fourdirections. A platform 10 cm in diameter was placed 2 cm below the water surface. Thelatency to escape onto the hidden platform was recorded by a computerized videoimaging analysis system (SLAC laboratory Animal Co. Ltd., Shanghai, China). The daybefore the test, each rat was allowed to swim freely for 120 s and locate the hiddenplatform. All rats received three trials per day for seven consecutive days, and thestarting quadrant was varied randomly over the trials. For each trial, rats were allowedto swim until they reach the platform and climbed onto it, subject to a 120-s cutoff. Anyanimal that failedwas guided to find the platform and allowed to stay on it for 3 s. On thelast day, the hidden platformwas removed, and the test trial was conducted. Each rat wasplaced in the water at the farthest point from the former location of the platform andallowed to swim for 120 s. The number of platform-crossing times was recorded.

2.6. Western blotting

Brain tissue was lysed in Western and IP lysis buffer on ice and homogenized.Lysates were centrifuged at 13,000 g for 10 min at 4°C. Protein concentrations of thesupernatants were determined with a BCA Protein Assay kit after separation. Afternormalization for total protein content, equal amounts of protein samples (20 μg perlane) were subjected to 10% (wt/vol) sodium dodecyl sulfate–polyacrylamide gelelectrophoresis; proteins were then transferred to nitrocellulose membranes andblocked with 5% (wt/vol) nonfat milk/Tris-buffered saline with Tween-20 (TBST)buffer for 1 h at room temperature. Membranes were incubated with primaryantibodies in 5% (wt/vol) milk/TBST at 4°C overnight. After washing with TBST three

times, membranes were incubated with horseradish peroxidase-conjugated secondaryantibody for 1 h at room temperature. Chemiluminescence detection was performedusing an ECL Western blotting detection kit and quantified by scanning densitometry.

2.7. Real-time PCR

Total RNA was isolated from the hippocampus using TRIzol reagent (Invitrogen)following the manufacturer's protocol. Reverse transcription from RNA into cDNA wasperformed using the PrimeScript RT-PCR Kit, followed by semiquantitative real-timePCR with specific primers. Glucocorticoid receptor (GR): CACCCATGATCCTGTCAGTG(forward), AAAGCCTCCCTCTGCTAACC (reverse); synaptophysin (SYP):AACAAAGGGCCTATGATGGA (forward), CCAGGTTCAGGAAGCCAAA (reverse);NMDA-R1: CTGCAACCCTCACTTTTGAG (forward), TGCAAAAGCCAGCTGCATCT (re-verse); NMDA-R2A: GACGGTCTTGGGATCTTAAC (forward), TGACCATGAATTGGTG-CAGG (reverse); NMDA-R2B: CAAGAACATGGCCAACCTGT (forward) ,GGTACACATTGCTGTCCTTC (reverse). Data were normalized to mRNA of actin as ahousekeeping gene and were analyzed by the 2−△△Ct method. Relative expression ofthe genes in treatment groups is expressed as percent of control.

2.8. Measurement of fetal serum corticosterone levels

Blood samples were centrifuged at 3000 rpm for 15 min, and serum was collectedand stored at −80°C. Serum corticosterone levels were determined with acommercially available corticosterone ELISA kit, following the manufacturer's protocol(RD Systems, Shanghai, China).

2.9. Statistics

All values are presented as the means±S.E.M. Latencies in the Morris water mazewere evaluated by two-way analysis of variance (ANOVA) for repeated measures.Other differences among groups were analyzed for significance using one-way ANOVA

Fig. 3. Effects of prenatal stress and HT supplementation on the expression of BDNF, SYP and GAP43. After male offspring were killed, protein and mRNA were extracted from thehippocampus. Expression of BDNF was determined by Western blotting (A: Western blot image; B: densitometric analysis of pro-BDNF; C: densitometric analysis of mature-BDNF).mRNA expression of SYP (D) and GAP43 (E) was determined by real-time PCR analysis. Values are mean values±S.E.M. n≥7, ⁎Pb.05, ⁎⁎Pb.01.

4 A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

followed by Dunnett's multiple comparison test. Values of Pb.05 were consideredstatistically significant.

3. Results

3.1. Offspring cognitive function

T-maze results showed that prenatal stress diminished spontane-ous alternation in both male and female offspring compared to thecontrol group (Fig. 1A, B). The number of arm entries was alsodecreased in male offspring of the stress groups (Fig. 1C) butunchanged in female rats (Fig. 1D). Maternal HT administrationsignificantly improved the number of arms entered in male offspringand increased the alternation behavior percentage in both male andfemale offspring (Fig. 1).

The Morris water maze results showed that prenatal stressinduced significant impairments in cognitive function in both maleand female offspring, and HT supplementation showed significantimprovements in the cognitive function of offspring (Fig. 2). As shownin Fig. 2A and B, two-way ANOVA (general linear model) repeatedmeasures revealed that HT produced a significant effect in males(F(3,60)=8.563, Pb.001) and females (F(3,50)=3.487, Pb.05,). LSDpost hoc analysis showed that stress induced poorer learning abilitythan controls in both male (Pb.001) and female (Pb.01) offspring.Both low-dose (male Pb.05; female P=.056) and high-dose (malePb.001; female Pb.01) HT decreased the escape latency timecompared with the stress-only group; there is a significant differencebetween the low-dose and high-dose groups in male (Pb.05) but notin female. For male, two-way ANOVA (general linear model)multivariate measures revealed a significant effect for HT at session1 [F(3, 60)=6.149, P=.001], session 2 [F(3, 60)=6.019, P=.001],

Fig. 4. Effects of prenatal stress and HT supplementation on the expression of NMDA receptors. After male offspring were killed, hippocampal mRNA was extracted. The mRNAexpression levels of the following NMDA receptors were determined by real-time PCR analysis: (A) NMDA-R1; (B) NMDA-R2A; (C) NMDA-R2B. Values are mean values±S.E.M. n≥7,⁎Pb.05, ⁎⁎Pb.01.

5A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

session 3 [F(3, 60)=3.564, P=.019] and session 4 [F(3,60)=6.586,P=.001]; for female, at session 3 [F(3, 50)=4.609, P=.006] andsession 5 [F(3, 50)=3.573, P=.020]. LSD post hoc showed significantdifferences at session 1–6 in male; at sessions 2, 3 and 5 in female.

As shown in Fig. 2C, stressed male offspring crossed the location ofthe platform many fewer times than control rats, consistent withprevious reports of cognitive dysfunction after prenatal stress.However, no significant difference was observed in female offspringbetween the stress and control groups at the probe test (Fig. 2D). HTsupplementation at both low and high doses showed significantimprovements in the cognitive function of male offspring, based onthe results of the Morris water maze probe test (Fig. 2A, C).

3.2. Offspring hippocampus neurogenesis

From the behavioral tests, we assumed that male offspring aremoresensitive than female offspring under prenatal stress. Therefore, wefocused on molecular changes in hippocampi of male offspring.Neurogenesis is responsible for populating the growing brain withneurons and for the development of cognitive function. We thusmeasured several important neuronalmarkers includingBDNF, SYP andgrowth-associated protein 43 (GAP43). As shown in Fig. 3, prenatalstress induced significant decreases in BDNF precursor protein(pro-BDNF) and mature BDNF protein levels in male offspring, andthese were efficiently improved by HT supplementation (Fig. 3A–C).mRNA expression levels of SYP and GAP43 in male offspring were

Fig. 5. Effects of prenatal stress and HT supplementation on the basal level of corticosteronecollected, and serum corticosterone levels were determined by ELISA (A). The mRNA expressValues are mean values±S.E.M. n≥7, ⁎Pb.05, ⁎⁎Pb.01.

significantly lower in the stress groups than in the control group(Fig. 3D, E), and maternal HT administration significantly increasedtheir expression compared to the stress group (Fig. 3D, E).

The glutamate system is closely related to neurogenesis andexcitatory neurotransmission in thebrain [26]. Previous studies showedthat prenatal stress significantly increased the glutamate concentrationin the hippocampi of offspring [27]. Here, we found that the glutamatereceptors N-methyl-D-aspartate receptor (NMDAR) R1, R2A and R2Bwere all affected by prenatal stress and HT supplementation. mRNAexpression levels of NMDAR-R1 and R2Bwere lower in the stress groupthan in the control group, and HT supplementation increased theirexpression (Fig. 4A, C). No significant difference was observed in theNMDAR-R2A level between the stress and control groups; however, HTsupplement significantly increased the NMDAR-R2A expression com-pared to the stress group (Fig. 4B).

3.3. GR and corticosterone

Fetal corticosterone levels were measured and showed nosignificant difference in the basal corticosterone levels of the controland stress groups in male offspring (Fig. 5A), similar to a previousreport by Henry et al. [28]. Meanwhile, expression of GR mRNA levelswas determined in the hippocampus. This revealed that GR mRNAwas reduced in the stress group compared to the control group inmale rats. Further, HT at both low and high doses increased the GRmRNA expression (Fig. 5B).

and the expression of GR. After the male offspring were killed, serum samples wereion levels of GR in male offspring hippocampus were determined by real-time PCR (B).

Fig. 6. Effects of prenatal stress and HT supplementation onmitochondrial content. After male offspring were killed, hippocampal proteins and DNAwere extracted. Protein expressionlevels of brain mitochondrial complex subunits were determined byWestern blot (A:Western blot image; B: densitometric analysis). The changes inmtDNA content were analyzed byreal-time PCR (C). Values are mean values±S.E.M. n≥7, ⁎Pb.05, ⁎⁎Pb.01.

6 A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

3.4. Mitochondrial function

As the key energy provider, mitochondria are closely related tovarious cellular activities including neurogenesis. Thus, we nextinvestigated mitochondrial content in male offspring hippocampus.As shown in Fig. 6, levels of mitochondrial complexes I and III weredecreased in the stress group compared to the control group (Fig. 6A,B). Supplementation with high-dose HT showed significant increasesin the levels of complex I, II, III, IV and V components compared to thestress group (Fig. 6A, B). Meanwhile, the mitochondrial DNA(mtDNA) copy number was decreased in the stress group comparedto the control group and increased by HT supplementation,suggesting that HT inhibited the mitochondrial loss induced byprenatal stress (Fig. 6C).

3.5. Oxidative damage and phase II enzymes

As we previously reported, prenatal stress can induce significantoxidative damage in the offspring's hippocampus. Here, we found thatthe increased protein oxidation indicated by the protein carbonylcontent was decreased in the HT supplemented groups (Fig. 7).Because HT is also a known activator of phase II enzymes to reduceoxidative stress, we examined the Nrf2 protein, a key regulator of thephase II defense system, and we found that HT could increase Nrf2expression and induce the expression of Nrf2 target gene HO-1(Fig. 8A–D). Meanwhile, prenatal stress-induced decrease of SOD2expression (Fig. 8E, F) and total SOD activity (Fig. 8G) wassignificantly improved after HT supplement. FOXO3, a knownregulator of SOD, was found to be decreased in the stress group and

increased in the HT-supplemented groups (Fig. 9). Meanwhile,FOXO1 was also found increased in the HT-supplemented groups,suggesting the possible involvement of FOXO family proteins in theregulation of phase II enzyme activity (Fig. 9).

4. Discussion

There is increasing evidence that prenatal stress can inducedamage to the brain structure and function and can result in mentaland behavioral abnormalities in offspring. The underlyingmechanismhas not been fully elucidated, but it has been reported that oxidativestress, mitochondrial dysfunction and disturbances in the function ofthe HPA axis are likely to be involved. Our present study firstdemonstrated that HT administration during pregnancy couldameliorate the impairments of learning and memory performanceinduced by maternal prenatal stress through the modulation ofmitochondrial populations and phase II enzymes.

Synaptic plasticity plays a central role in learning and memory[29]. In particular, it is closely associated with the memory lossmediated by stress to the hippocampus [30]. BDNF is known to boostneurogenesis and the long-term survival of newborn hippocampalneurons [31]. Downregulated BDNF expression may lead to reducedsynaptic plasticity and thus impair learning [32]. It has beensuggested that SYP, GAP43 and CREB are also involved in the effectsof BDNF on neuronal plasticity [33]. SYP is also required forneurotransmitter release, and loss of SYP in the hippocampus iscorrelated with the cognitive dysfunction seen in Alzheimer's disease[34]. In our study, prenatal stress decreased the expression of BDNF

Fig. 7. Effects of prenatal stress and HT supplementation on protein oxidation. Aftermale offspring were killed, hippocampal proteins were extracted. Protein oxidationlevel was detected by measuring the content of carbonyl protein via Western blotting(A: Western blot image, B: densitometric analysis). Values are mean values±S.E.M.n≥7, ⁎Pb.05, ⁎⁎Pb.01.

7A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

protein and the mRNA levels of SYP and GAP43 in the hippocampi ofmale offspring, all of which were significantly improved by HTsupplementation. Meanwhile, it has been reported that learningdeficits induced by maternal stress are associated with impairment ofNMDA receptor-mediated synaptic plasticity [35]. In the currentstudy, HT increased the expression of NMDA subunits NR1, NR2A andNR2B in prenatally stressed offspring. We thus propose that thepotential neuroprotective mechanism of HT may be associated withits ability to protect neurogenesis and modulate synaptic plasticity.

It is well known that glucocorticoid induces its downstream effectsby binding to the GR, and prenatal stress has been shown to reduce GRand/or mineralocorticoid receptor mRNA and protein expression inoffspring [36]. In the current study, the basal level of corticosterone

was not different among those four groups, which is consistent with aprevious study [28]. Reduced GR expression in the hippocampusmakes it difficult for the organism to regulate glucocorticoid secretionunder new stresses and enhances the organism's susceptibility tostress-related diseases. It was suggested that the decrease in GRexpression in hippocampus may not regulate HPA but may directlycause the functional impairment of learning andmemory in the stressresponse [36,37]. Mice with reduced GR expression showed adownregulation of BDNF protein content [38], whereas overexpres-sion of GR in mice results in BDNF overexpress in the hippocampus[39]. In our study, the decrease in GRmRNA levels after prenatal stressis consistent with previous findings, and HT treatment significantlyimproved the GR mRNA levels, and this may help to prevent theeffects of the stress response on cognitive function.

Oxidative stress and mitochondrial dysfunction have central rolesin neurodegenerative diseases [40,41]. In addition to neurotransmit-ter and neurotrophic factor signaling pathways, mitochondria play animportant role in the regulation of neuronal plasticity [42,43]. Ourprevious studies have demonstrated that immobilization or prenatalstress can result in oxidative damage and mitochondrial dysfunctionin the brain of the offspring [10,11,44]. In the present study, prenatalstress increased carbonyl protein content, which is a direct indicationof increased protein oxidation. Mitochondrial function and antioxi-dative phase II enzymes are known to regulate oxidative stress andprotect cell damage [45,46]. Here, prenatal stress induced a decline inmitochondrial biogenesis, evidenced by decreases in the mitochon-drial complex levels and the mtDNA copy number. Meanwhile, as thekey regulator of phase II enzymes, Nrf2 was found to be decreased byprenatal stress, and this may contribute to the decreased levels ofantioxidative enzymes and induce oxidative stress. HT was previouslyreported to be able to induce mitochondrial biogenesis and theactivation of phase II enzymes [19,47]. Consistently, offspring withmaternal HT supplement showed increased mitochondrial popula-tions and levels of antioxidative enzymes such as HO-1 and SOD2,which may significantly reduce the oxidative damage induced bystress and promote neuronal survival. In addition to Nrf2, a previousstudy also suggested that FOXO1 could regulate SOD2 expression tomodulate oxidative stress [48]. Further, FOXO genes, including bothFOXO1 and FOXO3, were found to be expressed in the hippocampusand to play vital roles in modulating brain function [49–51]. Anin vitro study suggested that HT could induce FOXO3 to preventincreased ROS levels in vascular endothelial cells [52]. Here, we foundincreased FOXO1 and FOXO3 expression in the HT-supplementedgroups, which might also contribute to the increased antioxidativeenzymes and reduced oxidative stress. Therefore, we propose that HTmay activate the FOXO/SOD and Nrf2/HO-1 pathways to prevent theoxidative stress induced by prenatal stress.

HT, the natural compound, has received increasing attention for itsbeneficial effects. It was reported that HT can be dose-dependentlyabsorbed and excreted in urine [53],with an estimatedhalf-life of 2.4 hin humans [54]. Rats HT plasmatic concentration can reach 1.22 μg/mlat 5min after 20mg/kgHT oral administration, and reached 1.91 μg/mlat 10min [55].Meanwhile, high-dose supplement of olive pulp extractrich in HT for 90 days had no adversely effects on rats mating, fertility,delivery and litter parameters [56]. Our previous study showed that25 mg/kg/day HT treatment had no adverse effect on inflammationand kidney function markers [20]. Also, toxicological evaluation of HTproposed a No Observed Adverse Effects level of 500 mg/kg/day [57].Therefore, previous studies have ensured that the doses we used haveno adverse effects on both mother rats and offspring. Also, recentreports indicated that the protective role of Mediterranean diet inpregnancy for the health of the mother and child is mainly due to theantioxidants supplied by components of this diet like olive oil [12,58].We thus proposed that HT would also show beneficial effects on brainfunction of normal offspring without stress. However, the detail

Fig. 8. Effects of prenatal stress and HT supplementation on phase II enzymes. After male offspring were killed, hippocampal proteins were extracted. The activation of the phase IIantioxidant system was detected by Western blot: Nrf2 (A: Western blot image, B: densitometric analysis); HO-1 (C: Western blot image, D: densitometric analysis); SOD2 (E:Western blot image, F: densitometric analysis). The activity of SOD2 was analyzed with hippocampus homogenates (F). Values are mean values±S.E.M. n≥7, ⁎Pb.05, ⁎⁎Pb.01.

8 A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

alternations and potential mechanisms still require furtherinvestigation.

In conclusion, our study demonstrates that prenatal stress canreducemitochondrial content and antioxidative capacity and increaseoxidative stress in the hippocampus of offspring, contributing todecreased neurogenesis and cognitive function. Maternal HT admin-istration successfully promotes cognitive function through modula-tion of mitochondrial content and phase II enzymes. This studyprovides a new candidate for nutritional intervention on stressmanagement. Hopefully, our study and proposed mechanism willhelp other investigators in this field.

Conflict of interest

There are no potential conflicts of interest relevant to this article.

Acknowledgments

This work was supported by the National Natural ScienceFoundation of China (Nos. 81201023 and 31370844), the 973Program (No. 2014CB548200), and the 985 and 211 Projects ofXi'an Jiaotong University.

Fig. 9. Effects of prenatal stress and HT supplementation on FOXO expression. Aftermale offspring were killed, hippocampal proteins were extracted. The expression levelsof FOXO1 and FOXO3 were detected by Western blot (A: Western blot image, B:densitometric analysis of FOXO1, C: densitometric analysis of FOXO3. Values are meanvalues±S.E.M. n≥7, ⁎Pb.05, ⁎⁎Pb.01.

9A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

References

[1] Huizink AC, Robles de Medina PG, Mulder EJ, Visser GH, Buitelaar JK. Stress duringpregnancy is associated with developmental outcome in infancy. J Child PsycholPsychiatry 2003;44:810–8.

[2] Markham JA, Koenig JI. Prenatal stress: role in psychotic and depressive diseases.Psychopharmacology 2011;214:89–106.

[3] Kinney DK. Prenatal stress and risk for schizophrenia. Int J Ment Health 2000;29:62–72.

[4] Laplante DP, Brunet A, Schmitz N, Ciampi A, King S. Project Ice Storm: prenatalmaternal stress affects cognitive and linguistic functioning in 5½-year-oldchildren. J Am Acad Child Adolesc Psychiatry 2008;47:1063–72.

[5] Nicoletto SF, Rinaldi A. In the womb's shadow: the theory of prenatalprogramming as the fetal origin of various adult diseases is increasinglysupported by a wealth of evidence. EMBO Rep 2011;12:30–4.

[6] Dancause KN, Laplante DP, Fraser S, Brunet A, Ciampi A, Schmitz N, et al. Prenatalexposure to a natural disaster increases risk for obesity in 5½-year-old children.Pediatr Res 2012;71:126–31.

[7] Skoluda N, Nater UM. Consequences of developmental stress in humans: prenatalstress. Adaptive andMaladaptive Aspects of Developmental Stress. Springer; 2013121–45.

[8] Lemaire V, Koehl M, Le Moal M, Abrous D. Prenatal stress produces learningdeficits associated with an inhibition of neurogenesis in the hippocampus. ProcNatl Acad Sci 2000;97:11032–7.

[9] Schmitz C, Rhodes M, Bludau M, Kaplan S, Ong P, Ueffing I, et al. Depression:reduced number of granule cells in the hippocampus of female, but not male, ratsdue to prenatal restraint stress. Mol Psychiatry 2002;7:810–3.

[10] Feng Z, Zou X, Jia H, Li X, Zhu Z, Liu X, et al. Maternal docosahexaenoic acid feedingprotects against impairment of learning and memory and oxidative stress inprenatally stressed rats: possible role of neuronal mitochondria metabolism.Antioxid Redox Signal 2012;16:275–89.

[11] Feng Z, Jia H, Li X, Bai Z, Liu Z, Sun L, et al. A milk-based wolfberry preparationprevents prenatal stress-induced cognitive impairment of offspring rats, andinhibits oxidative damage andmitochondrial dysfunction in vitro. NeurochemRes2010;35:702–11.

[12] Mariscal-Arcas M, Monteagudo C, Olea-Serrano F. Diet quality in pregnancy: afocus on requirements and the protective effects of the mediterranean diet. DietQuality. Springer; 2013 81–92.

[13] Al-Gubory KH. Maternal nutrition, oxidative stress and prenatal developmentaloutcomes. Studies on Women's Health. Springer; 2013 1–31.

[14] Luo C, Li Y, Wang H, Cui Y, Feng Z, Li H, et al. Hydroxytyrosol promotes superoxideproduction and defects in autophagy leading to anti-proliferation and apoptosison human prostate cancer cells. Curr Cancer Drug Targets 2013;13:625–39.

[15] Richard N, Arnold S, Hoeller U, Kilpert C, Wertz K, Schwager J. Hydroxytyrosol isthe major anti-inflammatory compound in aqueous olive extracts and impairscytokine and chemokine production in macrophages. Planta Med 2011;77:1890–7.

[16] González-Correa JA, Navas MD, Lopez-Villodres JA, Trujillo M, Espartero JL, De LaCruz JP. Neuroprotective effect of hydroxytyrosol and hydroxytyrosol acetate inrat brain slices subjected to hypoxia–reoxygenation. Neurosci Lett 2008;446:143–6.

[17] Schaffer S, Podstawa M, Visioli F, Bogani P, Müller WE, Eckert GP. Hydroxytyrosol-rich olive mill wastewater extract protects brain cells in vitro and ex vivo. J AgricFood Chem 2007;55:5043–9.

[18] Zhu L, Liu Z, Feng Z, Hao J, Shen W, Li X, et al. Hydroxytyrosol protects againstoxidative damage by simultaneous activation of mitochondrial biogenesis andphase II detoxifying enzyme systems in retinal pigment epithelial cells. J NutrBiochem 2010;21:1089–98.

[19] Zou X, Feng Z, Li Y, Wang Y, Wertz K, Weber P, et al. Stimulation of GSH synthesisto prevent oxidative stress-induced apoptosis by hydroxytyrosol in human retinalpigment epithelial cells: activation of Nrf2 and JNK-p62/SQSTM1 pathways. J NutrBiochem 2012;23:994–1006.

[20] Feng Z, Bai L, Yan J, Li Y, ShenW,Wang Y, et al. Mitochondrial dynamic remodelingin strenuous exercise-induced muscle and mitochondrial dysfunction: regulatoryeffects of hydroxytyrosol. Free Radic Biol Med 2011;50:1437–46.

[21] Morley-Fletcher S, Mairesse J, Maccari S. Behavioural and neuroendocrineconsequences of prenatal stress in rat. Adaptive and Maladaptive Aspects ofDevelopmental Stress. Springer; 2013 175–93.

[22] Cao K, Xu J, Zou X, Li Y, Chen C, Zheng A, et al. Hydroxytyrosol prevents diet-induced metabolic syndrome and attenuates mitochondrial abnormalities inobese mice. Free Radic Biol Med 2014;67:396–407.

[23] Deacon RM, Rawlins JNP. T-maze alternation in the rodent. Nat Protoc 2006;1:7–12.

[24] Fernandez F, Morishita W, Zuniga E, Nguyen J, Blank M, Malenka RC, et al.Pharmacotherapy for cognitive impairment in a mouse model of Down syndrome.Nat Neurosci 2007;10:411–3.

[25] Gué M, Bravard A, Meunier J, Veyrier R, Gaillet S, Recasens M, et al. Sex differencesin learning deficits induced by prenatal stress in juvenile rats. Behav Brain Res2004;150:149–57.

[26] Herlenius E, Lagercrantz H. Development of neurotransmitter systems duringcritical periods. Exp Neurol 2004;190:8–21.

[27] Jia N, Yang K, Sun Q, Cai Q, Li H, Cheng D, et al. Prenatal stress causes dendriticatrophy of pyramidal neurons in hippocampal CA3 region by glutamate inoffspring rats. Dev Neurobiol 2010;70:114–25.

[28] Henry C, Kabbaj M, Simon H, Moal M, Maccari S. Prenatal stress increases thehypothalamo‐pituitary‐adrenal axis response in young and adult rats. J Neu-roendocrinol 1994;6:341–5.

[29] Silva AJ. Molecular and cellular cognitive studies of the role of synaptic plasticityin memory. J Neurobiol 2003;54:224–37.

[30] Kim JJ, Diamond DM. The stressed hippocampus, synaptic plasticity and lostmemories. Nat Rev Neurosci 2002;3:453–62.

[31] Zuena AR, Mairesse J, Casolini P, Cinque C, Alemà GS, Morley-Fletcher S, et al.Prenatal restraint stress generates two distinct behavioral and neurochemicalprofiles in male and female rats. PLoS One 2008;3:e2170.

[32] Linnarsson S, Björklund A, Ernfors P. Learning deficit in BDNF mutant mice. Eur JNeurosci 1997;9:2581–7.

[33] Molteni R, Barnard R, Ying Z, Roberts C, Gomez-Pinilla F. A high-fat, refined sugardiet reduces hippocampal brain-derived neurotrophic factor, neuronal plasticity,and learning. Neuroscience 2002;112:803–14.

[34] Sze C-I, Troncoso JC, Kawas C, MOUTION P, Price DL, Martin LJ. Loss of thepresynaptic vesicle protein synaptophysin in hippocampus correlates withcognitive decline in Alzheimer disease. J Neuropathol Exp Neurol 1997;56:933–44.

[35] Son GH, Geum D, Chung S, Kim EJ, Jo J-H, Kim C-M, et al. Maternal stress produceslearning deficits associated with impairment of NMDA receptor-mediatedsynaptic plasticity. J Neurosci 2006;26:3309–18.

[36] Bingham BC, Rani S, Frazer A, Strong R, Morilak DA. Exogenous prenatalcorticosterone exposure mimics the effects of prenatal stress on adult brain

10 A. Zheng et al. / Journal of Nutritional Biochemistry xx (2014) xxx–xxx

stress response systems and fear extinction behavior. Psychoneuroendocrinology2013;38:2746–57.

[37] Welberg L, Seckl J, Holmes M. Prenatal glucocorticoid programming of braincorticosteroid receptors and corticotrophin-releasing hormone: possible impli-cations for behaviour. Neuroscience 2001;104:71–9.

[38] Ridder S, Chourbaji S, Hellweg R, Urani A, Zacher C, Schmid W, et al. Mice withgenetically altered glucocorticoid receptor expression show altered sensitivity forstress-induced depressive reactions. J Neurosci 2005;25:6243–50.

[39] Schulte-Herbrüggen O, Chourbaji S, Ridder S, Brandwein C, Gass P, Hörtnagl H,et al. Stress-resistant mice overexpressing glucocorticoid receptors displayenhanced BDNF in the amygdala and hippocampus with unchanged NGF andserotonergic function. Psychoneuroendocrinology 2006;31:1266–77.

[40] Lin MT, Beal MF. Mitochondrial dysfunction and oxidative stress in neurodegen-erative diseases. Nature 2006;443:787–95.

[41] Mancuso M, Coppede F, Migliore L, Siciliano G, Murri L. Mitochondrialdysfunction, oxidative stress and neurodegeneration. J Alzheimers Dis 2006;10:59–73.

[42] Mattson MP. Mitochondrial regulation of neuronal plasticity. Neurochem Res2007;32:707–15.

[43] Levy M, Faas GC, Saggau P, Craigen WJ, Sweatt JD. Mitochondrial regulation ofsynaptic plasticity in the hippocampus. J Biol Chem 2003;278:17727–34.

[44] Liu J, Wang X, Shigenaga M, Yeo H, Mori A, Ames B. Immobilization stress causesoxidative damage to lipid, protein, and DNA in the brain of rats. FASEB J 1996;10:1532–8.

[45] Zhu H, Jia Z, Strobl JS, Ehrich M, Misra HP, Li Y. Potent induction of total cellularandmitochondrial antioxidants and phase 2 enzymes by cruciferous sulforaphanein rat aortic smooth muscle cells: cytoprotection against oxidative andelectrophilic stress. Cardiovasc Toxicol 2008;8:115–25.

[46] Vincent AM, Kato K, McLean LL, Soules ME, Feldman EL. Sensory neurons andschwann cells respond to oxidative stress by increasing antioxidant defensemechanisms. Antioxid Redox Signal 2009;11:425–38.

[47] Hao J, Shen W, Yu G, Jia H, Li X, Feng Z, et al. Hydroxytyrosol promotesmitochondrial biogenesis and mitochondrial function in 3T3-L1 adipocytes. J NutrBiochem 2010;21:634–44.

[48] Higuchi M, Dusting GJ, Peshavariya H, Jiang F, Hsiao ST, Chan EC, et al.Differentiation of human adipose-derived stem cells into fat involves reactiveoxygen species and Forkhead box O1 mediated upregulation of antioxidantenzymes. Stem Cells Dev 2013;22:878–88.

[49] HoekmanMF, Jacobs FM, Smidt MP, Burbach JPH. Spatial and temporal expressionof FoxO transcription factors in the developing and adult murine brain. Gene ExprPatterns 2006;6:134–40.

[50] Polter A, Yang S, Zmijewska AA, van Groen T, Paik JH, Depinho RA, et al. Forkheadbox, class O transcription factors in brain: regulation and behavioral manifesta-tion. Biol Psychiatry 2009;65:150–9.

[51] Du Y, Zhang X, Ji H, Liu H, Li S, Li L. Probucol and atorvastatin in combinationprotect rat brains in MCAOmodel: upregulating Peroxiredoxin2, Foxo3a and Nrf2expression. Neurosci Lett 2012;509:110–5.

[52] Zrelli H, Matsuoka M, Kitazaki S, Zarrouk M, Miyazaki H. Hydroxytyrosol reducesintracellular reactive oxygen species levels in vascular endothelial cells byupregulating catalase expression through the AMPK–FOXO3a pathway. Eur JPharmacol 2011;660:275–82.

[53] Visioli F, Caruso D, Plasmati E, Patelli R, Mulinacci N, Romani A, et al.Hydroxytyrosol, as a component of olive mill waste water, is dose-dependentlyabsorbed and increases the antioxidant capacity of rat plasma. Free Radic Res2001;34:301–5.

[54] Miro-Casas E, Covas MI, Farre M, Fito M, Ortuno J, Weinbrenner T, et al.Hydroxytyrosol disposition in humans. Clin Chem 2003;49:945–52.

[55] Ruiz-Gutierrez V, Juan ME, Cert A, Planas JM. Determination of hydroxytyrosol inplasma by HPLC. Anal Chem 2000;72:4458–61.

[56] Christian MS, Sharper VA, Hoberman AM, Seng JE, Fu L, Covell D, et al. The toxicityprofile of hydrolyzed aqueous olive pulp extract. Drug Chem Toxicol 2004;27:309–30.

[57] Aunon-Calles D, Canut L, Visioli F. Toxicological evaluation of pure hydroxytyr-osol. Food Chem Toxicol 2013;55:498–504.

[58] Castro‐Rodriguez JA, Garcia‐Marcos L, Sanchez‐Solis M, Pérez‐Fernández V,Martinez‐Torres A, Mallol J. Olive oil during pregnancy is associated with reducedwheezing during the first year of life of the offspring. Pediatr Pulmonol 2010;45:395–402.