ORIGINAL RESEARCH published: 19 December 2019 doi: 10.3389/fnagi.2019.00346 Ovarian Hormone-Dependent Effects of Dietary Lipids on APP/PS1 Mouse Brain Jose Luis Herrera 1 , Lara Ordoñez-Gutierrez 2,3 , Gemma Fabrias 4 , Josefina Casas 4 , Araceli Morales 1 , Guadalberto Hernandez 1 , Nieves G. Acosta 5 , Covadonga Rodriguez 1 , Luis Prieto-Valiente 6 , Luis M. Garcia-Segura 7,8* , Francisco G. Wandosell 2,3* and Rafael Alonso 1* 1 Departamento de Ciencias Médicas Básicas, Instituto de Tecnologías Biomédicas–Centro de Investigaciones Biomédicas de Canarias, Universidad de La Laguna, San Cristóbal de La Laguna, Spain, 2 Centro de Biología Molecular “Severo Ochoa” (CSIC-UAM), Universidad Autónoma de Madrid, Madrid, Spain, 3 Centro de Investigación Biomédica en Red de Enfermedades Neurodegenerativas, Madrid, Spain, 4 Instituto de Química Avanzada de Cataluña (IQAC-CSIC), Barcelona, Spain, 5 Departamento de Biología Animal, Edafología y Geología, Instituto de Tecnologías Biomédicas–Centro de Investigaciones Biomédicas de Canarias, Universidad de La Laguna, San Cristóbal de La Laguna, Spain, 6 Unidad de Estadística Médica, Universidad Católica de Murcia, Murcia, Spain, 7 Instituto Cajal, CSIC, Madrid, Spain, 8 Centro de Investigación Biomédica en Red de Fragilidad y Envejecimiento Saludable, Madrid, Spain The formation of senile plaques through amyloid-β peptide (Aβ) aggregation is a hallmark of Alzheimer’s disease (AD). Irrespective of its actual role in the synaptic alterations and Edited by: cognitive impairment associated with AD, different therapeutic approaches have been George E. Barreto, proposed to reduce plaque formation. In rodents, daily intake of omega-3 (n-3) long- University of Limerick, Ireland chain polyunsaturated fatty acids (LC-PUFAs) is required for neural development, and Reviewed by: Ann Van Der Jeugd, there is experimental and epidemiological evidence that their inclusion in the diet has KU Leuven, Belgium positive effects on several neurodegenerative diseases. Similarly, estradiol appears to Marco Fidel Avila-Rodriguez, reduce senile plaque formation in primary mouse cell cultures, human cortical neurons Universidad del Tolima, Colombia Ramesh Kandimalla, and mouse AD models, and it prevents Aβ toxicity in neural cell lines. We previously Texas Tech University Health showed that differences in dietary n-6/n-3 LC-PUFAs ratios modify the lipid composition Sciences Center, United States in the cerebral cortex of female mice and the levels of amyloid precursor protein (APP) *Correspondence: Luis M. Garcia-Segura in the brain. These effects depended in part on the presence of circulating estradiol. lmgs@cajal.csic.es Here we explored whether this potentially synergistic action between diet and ovarian Francisco G. Wandosell hormones may influence the progression of amyloidosis in an AD mouse model. Our fwandosell@cbm.csic.es Rafael Alonso results show that a diet with high n-3 LC-PUFA content, especially DHA (22:6n-3), ralonsosolis@gmail.com reduces the hippocampal accumulation of Aβ1−40 , but not amyloid Aβ1−42 in female APPswe/PS1 E9A mice, an effect that was counteracted by the loss of the ovaries Received: 03 July 2019 Accepted: 26 November 2019 and that depended on circulating estradiol. In addition, this interaction between dietary Published: 19 December 2019 lipids and ovarian function also affects the composition of the brain lipidome as well Citation: as the expression of certain neuronal signaling and synaptic proteins. These findings Herrera JL, Ordoñez-Gutierrez L, Fabrias G, Casas J, Morales A, provide new insights into how ovarian hormones and dietary composition affect the brain Hernandez G, Acosta NG, lipidome and amyloid burden. Furthermore, they strongly suggest that when designing Rodriguez C, Prieto-Valiente L, dietary or pharmacological strategies to combat human neurodegenerative diseases, Garcia-Segura LM, Wandosell FG and Alonso R (2019) Ovarian hormonal and metabolic status should be specifically taken into consideration as it may Hormone-Dependent Effects affect the therapeutic response. of Dietary Lipids on APP/PS1 Mouse Brain. Front. Aging Neurosci. 11:346. Keywords: cerebral cortex lipidome, long-chain polyunsaturated fatty acids, docosahexaenoic acid, doi: 10.3389/fnagi.2019.00346 sphingolipids, ovarian hormones, synaptic proteins, amyloid, estradiol Frontiers in Aging Neuroscience | www.frontiersin.org 1 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function INTRODUCTION Morris et al., 2005; Scarmeas et al., 2006; Cederholm and Palmblad, 2010; Shinto et al., 2014). Furthermore, the results Alzheimer’s Disease (AD) is characterized by the presence of a randomized, double-blind, placebo-controlled clinical trial of senile plaques and neurofibrillary tangles in the brain, suggested that administration of n-3 LC-PUFAs may prevent which ultimately lead to progressive neuronal dysfunction and AD, although this does not represent a treatment for the disease dramatically impaired cognitive performance (Duyckaerts, 2003). (Freund-Levi et al., 2006). It is accepted that the formation of senile plaques is a consequence One issue that has not been completely addressed in all these of amyloid-β (Aβ) polymerization, a peptide derived from the earlier studies is that of “gender.” Human population studies Amyloid Precursor Protein (APP: Selkoe, 2002). According to have shown a clear increase in the risk of suffering AD and the amyloidogenic theory, Aβ-induced toxicity leads to a cascade other dementias in women after menopause (Manly et al., 2000), of events that ultimately produce oxidative stress, mitochondrial suggesting a potential neuroprotective role for ovarian hormones. dysfunction, and synaptic impairment (Hardy and Selkoe, 2002; In rodents, the ovarian hormone 17β-estradiol (estradiol) Gibson and Huang, 2005). A small proportion of AD cases regulates brain plasticity and cognition (reviewed in Barha and have a genetic origin, classified as Familiar Alzheimer Disease Galea, 2010), and support for its role as a neuroprotectant has (FAD), caused by mutations in the APP (Mullan et al., 1992), been obtained in cellular models of Aβ-induced neurotoxicity Presenilin 1 (Sherrington et al., 1996) or Presenilin 2 (Levy-Lahad (Xu et al., 1998; Marin et al., 2003; Guerra et al., 2004). In et al., 1995) genes. By contrast, more than 95% of AD cases addition, both inhibition and knockout of aromatase, the enzyme are sporadic, which implies some allelic variation in the normal that converts testosterone to estradiol, make AD pathology in population, and that aging and environmental or dietary factors animal models worse (McCullough et al., 2003; Overk et al., influence this pathology. Indeed, although the relationship with 2012). Furthermore, aromatase gene polymorphisms may affect synaptic dysfunction remains unclear, AD has been associated the risk of AD in humans, particularly in women (Iivonen et al., with metabolic disorders, such as diabetes, hypercholesterolemia 2004; Huang and Poduslo, 2006; Janicki et al., 2013; Medway or hyperlipidemia (Ricciarelli et al., 2012; Park et al., 2013; et al., 2014). Taken together, all these data strongly suggest Dar et al., 2014; Baglietto-Vargas et al., 2016; Kandimalla et al., that the presence of circulating gonadal hormones may have 2017; Pugazhenthi et al., 2017; Salas and De Strooper, 2019). In a neuroprotective effect against several neurological diseases, this respect, brain insulin resistance and IRS-1 dysfunction has although this phenomenon has not been adequately considered been detected in AD patients (Talbot et al., 2012), and diabetes in the context of dietary interventions. drugs represent promising therapies for AD (Ferreira et al., 2018; We previously found that different concentrations of n-6 and Boccardi et al., 2019). n-3 LC-PUFAs in the diet are associated with differences in the Some long-chain polyunsaturated fatty acids (LC-PUFAs) total fatty acids (FAs), sphingolipids, and gangliosides in the like docosahexanoic acid (DHA, 22:6n-3) and pro-resolving female mouse cerebral cortex. Interestingly, these changes in lipid mediators (SPMs) like neuroprotectin D1 (NPD-1) or the brain lipidome composition are partially dependent on the resolving D5 (RvD5), modulate signaling processes mediated by presence or absence of circulating ovarian hormones (Herrera a variety of receptors, such as lipoxin A4 receptor, G protein- et al., 2018). Furthermore, we also found that diet and ovarian coupled receptor 32, chemerin receptor 23, and leukotriene function interact and affect the expression of several neuronal B4 receptor, which regulate cell survival, neuroinflammation, proteins as well as the APP levels in the brain. Here, we set neurotransmission, and cognitive performance (Bannenberg and out to determine whether the intake of several of the most Serhan, 2010; Orr et al., 2013; Serhan and Chiang, 2013; Serhan, common n-3 LC-PUFAs, such as 20:5n-3 (EPA) and DHA, 2014). In humans, diets enriched in LC-PUFAs have been modifies the brain lipidome, the expression of neuronal proteins, associated with a lower incidence of dementia and neurological and the levels of Aβ. In addition, we analyzed whether the effects disorders (Mazza et al., 2007), while in rodents, n-3 LC- of the dietary interventions were influenced by the presence PUFA poor diets, such as those with limited DHA, have or absence of circulating estradiol. To this end, we used the been associated with cognitive impairment (Ikemoto et al., APPswe/PS11E9 (APP/PS1) double transgenic mouse model 2001; Catalan et al., 2002). By contrast, high levels of DHA of cerebral amyloidosis, in which increased Aβ levels develop reduce the total Aβ and senile plaque formation, protecting AD by 6–8 months of age (Trinchese et al., 2004) as well as early mouse models from neuronal pathologies (Calon et al., 2004; synaptic defects (Viana da Silva et al., 2016). These animals Lim et al., 2005; Green et al., 2007; Perez et al., 2010) and were fed two experimental diets: (a) a diet with a particularly improving spatial memory and cognition (Hooijmans et al., high content of 18:2n-6 PUFAs and a high n-6/n-3 PUFA 2009). Aβ peptides are generated from the transmembrane APP ratio, also lacking DHA and other common n-3 and n-6 LC- through the sequential activity of two membrane associated PUFAs (DI); or (b) a diet with a lower n-6/n-3 PUFA ratio, proteases, BACE and γ-secretase. Thus, such cleavage is particularly enriched in n-3 LC-PUFAs, DHA, and EPA (DII). likely to be strongly influenced by the lipidic milieu and The effects of these diets on the brain lipidome and Aβ levels membrane trafficking (Hicks et al., 2012; Das et al., 2013, 2016). were compared with that of standard laboratory food (SF) that Several epidemiological studies suggest that dietary composition lacks LC-PUFAs but contains the LC-PUFA precursors linoleic influences brain function, and specifically, that certain LC-PUFAs acid (18:2n-6) and linolenic acid (18:3n-3). At 90 days of age, exert protective effects in neurodegenerative disorders (Kalmijn all the mice were divided into three groups: (i) Sham-operated et al., 1997; Barberger-Gateau et al., 2002; Engelhart et al., 2002; controls; (ii) Ovariectomized-placebo treated (OVX); and (iii) Frontiers in Aging Neuroscience | www.frontiersin.org 2 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function Ovariectomized-estradiol treated (OVX-E). After 3 months Genotyping under these dietary and hormonal regimes, all mice were The animal’s genotype was confirmed using three primers: one sacrificed and their brains were dissected out and analyzed to antisense primer matching a sequence within PrP (50 GTG GAT determine the amyloid burden, lipidome composition, and key ACC CCC TCC CCC AGC CTA GAC C30 ); one sense primer synaptic and survival signaling proteins. Our findings support specific for the transgene (PS1, 50 CAG GTG GTG GAG CAA three major conclusions. First, diets containing different n-6 GAT G30 ; APP, 50 CCG AGA TCT CTG AAG TGA AGA TGG and n-3 PUFA ratios and LC-PUFA levels modify the cerebral ATG30 ); and a second sense primer specific for the genomic PrP cortex lipidome in female APP/PS1 mice distinctly, although (50 : CCT CTT TGT GAC TAT GTG GAC TGA TGT CGG30 ). the effects are not identical to those previously observed in wild type (WT) animals. Second, a diet with high levels of Ovariectomy and Estradiol Treatment DHA and other LC-PUFAs reduces the hippocampal levels of Mice were anesthetized with inhaled isoflurane (2 ± 0.5%) after amyloid Aβ1–40 in this AD model. Third, the effects of diet receiving an analgesic injection (buprenorphine hydrochloride, on the brain lipidome, amyloid burden and the expression of Buprex) and then ovariectomized bilaterally at 90 ± 1 days neuronal proteins are partially dependent on normal ovarian of age through a 1 cm dorsal incision that was then closed function. Together, these results reveal an interaction of diet with surgical clips (Table 1). The day after ovariectomy, mice and ovarian hormones that affects Aβ accumulation, lipidome received a 3 mm pellet containing 0.05 mg estradiol on a 90- composition, and neuronal protein expression in the brain of a day time-release or a placebo (Innovative Research of America, mouse model of AD. This new information may be useful in the Sarasota, FL, United States). The hormonal dose was chosen from search for novel therapeutic interventions to either prevent or preliminary experiments in which the levels of estradiol in 90- treat neurodegeneration in aging women. day-old treated animals were 7.2 ± 4.5 pg/ml (mean ± SEM), as determined by radioimmunoassay (RIA) (Architect System, ref #B7K720, Abbot, Germany). These levels are within the MATERIALS AND METHODS range of those observed in cycling animals of the same age at proestrus (20.9 ± 6.0 pg/ml) and estrus (1.8 ± 4.5 pg/ml). The Animals and Husbandry pellets were implanted subcutaneously in the subscapular region Specific pathogen-free mice of two inbred strains were used using a trocar, following the manufacturer’s instructions, and no in this study. Male APPswe/PS11E9 (APP/PS1) and female inflammation around the implantation area was observed after C57BL/6J mice (Mus musculus) were purchased from Jackson 3 days. Mice were then housed according to the different types of Laboratories and Charles River Laboratories, respectively. diet and hormone treatment until the end-point, 90 days later. Animals were housed under conditions of constant temperature (22 ± 2◦ C) and humidity (50 ± 5%), and on a 12-h light- Diets dark cycle with free access to SF (A03/R03, SAFE-Panlab) Two specific experimental diets were used in these experiments and tap water. Crosses from both strains were genotyped by (Table 2). The high n-6/n-3 ratio diet (DI) contained sunflower polymerase chain reaction (PCR) using DNA from tail biopsies oil and consequently, it was particularly abundant in linoleic acid (Jankowsky et al., 2001). Female APP/PS1 mice were selected for (18:2n-6) but poor in α-linolenic acid (18:n-3), with undetectable the experiments and housed (10 per cage) in cages fitted with levels of EPA (20:5n-3) and DHA (22:6n-3). The low n-6/n- microbarrier filter tops. 1 month after birth the SF was gradually 3 ratio diet (DII) had the same basic composition but it was replaced with the experimental diets (see below) at a rate of 25% specifically supplemented with extra EPA and DHA (7 g/kg), per week. Thus, all mice were fed 100% experimental diets from added as a lipid source in the form of fish oil to give a particularly the age of 2 months until the day of sacrifice (Table 1). The high DHA content. These diets were designed at the “Instituto protocols used in this study were all approved by the Animal Care de Nutrición y Tecnología de los Alimentos” of the University and Use Committee at University of La Laguna. of Granada (Spain), and produced by Mucedola (Mucedola srl, Milano, Italy). Both these diets and the SF were subsequently analyzed to determine the final percentage and absolute quantity TABLE 1 | Timeline of animal husbandry, feeding and experimental manipulations. (g/kg) of each FA. For simplicity, high and low n-6/n-3 PUFA ratio diets are referred to in the text and graphics as DI and DII, Months Animal manipulations respectively. Mice were fed these diets ad libitum for 90 days until 0 After birth mice were housed under constant temperature the day of sacrifice. (22 ± 2◦ C) and humidity (50 ± 5%) and a 12:12 h light-dark cycle, with free access to standard laboratory food Tissue Processing and Sample 1 From day 30 of age, standard laboratory food was gradually Preparation replaced with the experimental diets at a rate of 25% per week. Ninety days after ovariectomy or sham operation, the mice This dietary regime was maintained until the time of sacrifice at the age of 6 months. were sacrificed using CO2 and their brain was collected. For 3 At 90 days of age, mice were bilaterally ovariectomized or ELISAs, hippocampal tissue was homogenized in eight volumes sham-operated and implanted with pellets containing 90-day timed of ice-cold guanidine buffer containing 5 mol/L guanidine HCl, released 17β-estradiol (0.05 mg) or vehicle 50 mmol/L Tris HCl [pH 8]. Homogenates were then mixed 6 All mice were sacrificed at the age of 6 months for 3 h at room temperature and stored at −20◦ C. Cerebral Frontiers in Aging Neuroscience | www.frontiersin.org 3 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function TABLE 2 | Main fatty acid (FA) composition of experimental diets (DI and DII) and Dietary FA profiles (g FA/kg diet fresh weight) were obtained standard food (SF). by acid-catalyzed transmethylation of the lipid fractions Fatty acids SF DI DII followed by GC-MS (Gas Chromatography-Mass Spectrometry: Fabelo et al., 2014). Cerebral cortex lipid fractions were C 14: 0 0.05 ± 0.00 0.21 ± 0.01 1.50 ± 0.00 subjected to further analysis of the FAs and complex lipids as C 16: 0 4.13 ± 0.14 4.68 ± 0.31 6.33 ± 0.04 described previously (Cingolani et al., 2014). Sphingolipids were C 16:1 n-7 0.09 ± 0.01 0.17 ± 0.01 1.55 ± 0.01 analyzed by HPLC-MS using 0.2 nmol of C17-sphinganine, C 18:0 0.75 ± 0.04 1.55 ± 0.08 1.77 ± 0.01 N-dodecanoylsphingosine, N-dodecanoylglucosylsphingosine, C 18:1 n-9 13.15 ± 0.35 12.04 ± 0.81 6.15 ± 0.05 and N-dodecanoyl sphingosylphosphorylcholine as internal C 18:1 n-7 0.34 ± 0.06 0.59 ± 0.07 0.98 ± 0.02 standards. Sphingolipids were annotated as C 18:2 n-6 10.63 ± 0.78 28.84 ± 0.42 16.23 ± 0.00 : . If the sphingoid base residue was dihydrosphingosine, C 18:4 n-3 0.00 ± 0.00 0.00 ± 0.00 0.66 ± 0.01 the lipid class contained a ð prefix. In most cases, the final tissue C 20:0 0.11 ± 0.01 0.18 ± 0.00 0.24 ± 0.00 data were given as pmol/mg of protein, except in the case of C 20:1 n-9 0.18 ± 0.00 0.12 ± 0.00 1.09 ± 0.07 the total FA that was represented as pmol equivalents per mg of C 20:4 n-6 (ARA) 0.00 ± 0.00 0.00 ± 0.00 0.20 ± 0.00 protein with respect to C12 ceramide. For statistical analysis, the C 20:4 n-3 0.00 ± 0.00 0.00 ± 0.00 0.24 ± 0.01 brain levels of specific LC-PUFAs were shown as the percentage C 20:5 n-3 (EPA) 0.00 ± 0.00 0.00 ± 0.00 2.29 ± 0.02 of the total FA. C 22:0 0.18 ± 0.01 0.10 ± 0.01 0.13 ± 0.02 C 22:1 n-11 0.05 ± 0.02 0.10 ± 0.06 0.74 ± 0.04 Amyloid Aβ Measurement C 22: 5 n-6 (DPA) 0.00 ± 0.00 0.00 ± 0.00 0.11 ± 0.01 Commercial ELISA kits were used to measure the Aβ1–40 and C 22: 5 n-3 0.05 ± 0.00 0.06 ± 0.03 0.43 ± 0.01 Aβ1–42 (Aβ1 –40 human ELISA kit, KHB 3482 and Aβ1 –42 C 22: 6 n-3 (DHA) 0.00 ± 0.00 0.00 ± 0.00 3.47 ± 0.01 human ELISA kit KHB 3442, respectively; Invitrogen, Thermo TOTALS Fisher Scientific, Waltham, MA, United States). Briefly, brain Saturates 5.42 ± 0.16 6.83 ± 0.37 10.41 ± 0.06 homogenates were diluted 1:50 in PBS-Tween-BSA buffer (0.03% Monoenes 14.00 ± 0.39 13.41 ± 1.03 11.24 ± 0.09 Tween-20, 5% BSA in PBS) before centrifuging at 4◦ C for PUFAs 11.69 ± 0.91 29.05 ± 0.47 24.07 ± 0.08 20 min at 16,000 × g. Each assay was run in duplicate and the n-9 13.47 ± 0.33 12.35 ± 0.86 7.85 ± 0.10 plate absorbance was measured at 450 nm using an Opsys MR n-6 10.63 ± 0.78 28.84 ± 0.42 16.55 ± 0.09 microplate reader (Dynex Technologies). n-3 1.06 ± 0.13 0.22 ± 0.06 7.52 ± 0.01 n-3 LC-PUFAs 0.05 ± 0.00 0.06 ± 0.03 6.43 ± 0.02 Western Blot Analysis n-6 LC-PUFAs 0.00 ± 0.00 0.00 ± 0.00 0.32 ± 0.01 Cerebral cortex protein lysates (20 µg) were resolved by SDS- RATIOS PAGE, using 8 or 12% polyacrylamide gels, depending on the n-3/n-6 0.10 ± 0.01 0.01 ± 0.00 0.45 ± 0.00 molecular weight of the studied proteins, and transferred onto n-6/n-3 10.04 ± 0.52 138.37 ± 33.71 2.20 ± 0.02 nitrocellulose (Whatman) or PVDF (Millipore) membranes, Total FA content 31.53 ± 1.67 49.84 ± 2.10 46.43 ± 0.45 that were subsequently blocked by incubation in 10% non-fat % Lipids (fresh wt) 5.48 ± 0.11 6.45 ± 0.55 6.31 ± 0.00 milk for 1 h at room temperature and incubated overnight % Moisture 9.79 ± 0.09 8.65 ± 0.27 7.85 ± 0.00 at 4◦ C with the appropriate primary antibodies (Table 3) % Lipids (dry wt) 6.07 ± 0.12 7.06 ± 0.59 6.85 ± 0.00 in the conditions described previously (Herrera et al., 2018). Data are mean ± SD of g/kg of fresh weight. Totals include some minor Antibody binding was detected with Supersignal (Thermo components not shown. The most common LC-PUFAs mentioned in the text are bold marked. Fisher Scientific) using β-actin as the internal loading control. Densitometric analysis was performed using a GS-800 Calibrated cortex tissue for Western blotting was homogenized in three Densitometer (Bio-Rad) and the raw data obtained are presented volumes of ice-cold lysis buffer containing 20 mM HEPES, in Supplementary Table 1. 100 mM NaCl, 100 mM NaF, 1 mM Na3 VO4 , 5 mM EDTA, 1% Triton X-100, 1 µM Okadaic acid (Calbiochem) and a protease Statistical Analysis inhibitor cocktail (Roche Diagnostic). The homogenates were To address whether different dietary n-6/n-3 ratios of LC-PUFA processed as described previously (Herrera et al., 2018) and all modify the brain lipidome in intact female APP/PS1 mice, one- samples were probed in 3–4 Western blots. For lipid analysis, way ANOVA was followed by linear contrasts, the diet being the the tissues were homogenized at a concentration of 5 mg/ml experimental factor with three levels (SF, DI, and DII). Although in phosphate-buffered saline (PBS) with 0.01% 3,5-Di-tert-4- the actual values of the total FAs, Ceramides, dh-Ceramides, butylhydroxytoluene (BHT) as an antioxidant, and they were Sphingomyelins, and dh-Sphingomyelins are represented in the processed as described below. corresponding figure, due to the extremely large standard errors, a logarithmic transformation was used to perform the statistical Lipid Analysis analysis. For the general ANOVA, only a p value < 0.05 was Lipids were extracted from dietary samples and cerebral tissue considered to reflect a significant effect of dietary composition. using a modification of Folch’s method (Folch et al., 1957). To avoid undesirable effects of multi-testing due to significance Frontiers in Aging Neuroscience | www.frontiersin.org 4 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function TABLE 3 | Antibodies used in Western blot analysis of brain signaling and synaptic proteins. Antibody Host Dilution References PSD95 rabbit 1:1,000 #3450 Cell Signaling, United States PI3K-p85 mouse 1:1,000 #06-195 Upstate (Millipore), United States Phospho-synapsin (Ser9) rabbit 1:1,000 #2311 Cell Signaling, United States Synapsin rabbit 1:1,000 #2312 Cell Signaling, United States Akt (PKB) rabbit 1:1,000 #9272 Cell Signaling GSK-3α/β mouse 1:1,000 44-610 Invitrogen, United States Phospho-GSK-3 (Ser21/9) rabbit 1:1,000 #9331 Cell Signaling, United States GFAP rabbit 1:1,000 G5601 Promega, United States Synaptophysin (SY38) mouse 1:20,000 10701 Progen, United Kingdom β-Actin mouse 1:10,000 A5441 Sigma-Aldrich, United States Goat anti-mouse IgG-HRP goat 1:5,000 Sc-2005 Santa Cruz Biotech, Germany Goat anti-rabbit IgG-HRP goat 1:5,000 Sc-2004 Santa Cruz Biotech, Germany inflation, no more than two contrasts were tested for each was not significantly higher than in the SF (31.53). However, the variable. Thus, the effect of each experimental diet (DI or DII) three diets did differ in terms of the content of several relevant on each specific lipid class was compared to that of the SF. lipid species, such as the LC-PUFAs C20:4n-6 (ARA), C20:5n-3 To analyze the effect of diet and the potential influence of (EPA), C22:5n-6 (DPA), or C22:6n-3 (DHA). Neither SF nor DI hormonal status on the hippocampal levels of Aβ1–40 and Aβ1– presented measurable traces of these dietary constituents while 42 and on protein expression in the cerebral cortex, a 2 × 3 they were all present in DII especially EPA and in particular, factorial analysis was designed. Again, the two levels of the DHA. Consequently, there were dramatic differences in the dietary factors were DI and DII, and the three levels of the dietary ratios of n-3/n-6 (SF, 0.10; DI, 0.01; DII: 0.45) and n-6/n- gonadal factor were normal ovarian function (SHAM) and the 3 PUFAs (SF, 10.04; DI, 138.37; DII: 2.20). In summary, the main presence (OVX-E) or absence (OVX) of circulating estradiol dietary differences were: a) DI had a n-6/n-3 ratio 13–14 times in ovariectomized mice. After a general ANOVA analysis, we higher than that of SF and 59 times higher than that of DII; b) performed no more than three contrasts to answer two specific DII had a n-6/n-3 ratio 4–5 times lower than that of SF; and c) DI questions: (i) for each diet, DI or DII, did the absence of the had no traces of either DHA or EPA, while DII contained relevant ovaries induce a significant effect; (ii) whether the presence of amounts of both n-3 LC-PUFAs (Table 2). circulating estradiol could counteract any effect of ovariectomy; and (iii), whether the effect of ovariectomy and estradiol could be considered different when one diet was maintained as opposed Effect of the Different Diets on the Brain to the effect of the other. In addition, to estimate the effects Lipidome of Female APP/PS1 Mouse and interactions between different factors, we run a saturated According to Their Gonadal Status regression model that included the dichotomous variable diet and In terms of the brain lipidome, the effects of the experimental the polytomous variable hormone condition, with three levels diets were compared with the SF (Figure 1). In intact sham- (Armitage et al., 2004). The latter enters the model as two dummy operated female APP/PS1 mice, the diets with different ratios variables, the first with a value of “1” in ovariectomized animals of n-6/n-3 PUFAs and n-3 LC-PUFA induced distinct changes receiving placebo containing pellets (OVX), and the second with in the cerebral cortex lipidome depending on the lipid class a value of “1” in ovariectomized mice treated with estradiol analyzed. Animals fed either DI or DII had significantly more (OVX-E). Intact sham-operated controls had a value of “0” for total FAs than animals fed SF, 10- to 20-fold more on average each dummy variable. This model provides us with confidence (Figure 1A, p = 0.0005). DI also induced a significant increase intervals and tests the effects of the three diets conditioned to in the relative levels of ARA (Figure 1B, p = 0.0009) and DPA the hormonal status and the interaction between the two factors. (Figure 1D, p = 0.0005) compared to SF, as well as a significant This means four comparisons were planned “a priori” in the reduction in DHA (Figure 1E, p = 0.001) and in the DHA/DPA experimental design. ratio (Figure 1F, p = 0.05). By contrast, DII caused a significant decrease in DPA (Figure 1D, p = 0.02) and a significant increase in the DHA/DPA ratio (Figure 1F, p = 0.005). Conversely, RESULTS both the DI and DII diets significantly increased the total dh- Ceramides (Figure 1H, p = 0.02 and p = 0.002, respectively), The Composition of the Diets while DII produced a significant elevation in the levels of total The composition of the experimental diets (DI and DII) was Sphingomyelins (Figure 1I, p = 0.04) and dh-Sphingomyelins compared with that of the SF and the three diets did not differ (Figure 1J, p = 0.007). significantly in their total lipid dry weight, ranging from 6.1% in In addition to the differential effects of the diets, we also SF to 7.1% in DI and 6.9% in DII. The total FA content (g/kg studied the influence of gonadal status on dietary-dependent fresh weight) was also similar in DI (49.84) and DII (46.43), and it brain lipidome remodeling. As such, we compared the differences Frontiers in Aging Neuroscience | www.frontiersin.org 5 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function between the effects of DI and DII on the cerebral cortex lipidome in OVX and OVX-E mice. In ovariectomized APP/PS1 mice fed DI, estradiol administration increased the total FAs (Figure 2A, p = 0.02), the relative levels of DPA (Figure 2B, p = 0.0001), and the total levels of Ceramides (Figure 2E, p = 0.02), dh-Ceramides (Figure 2F, p = 0.01), Sphingomyelins (Figure 2G, p = 0.05), and dh-Sphingomyelins (Figure 2H, p = 0.05). In ovariectomized mice fed DII, the only relevant effect of estradiol administration was the elevation of the DHA/DPA ratio relative to the animals receiving the placebo (Figure 2D, p = 0.009). We also compared the differential effect of estradiol in the animals fed DI or DII using linear saturated models. The influence of estradiol on brain lipidome remodeling depended on each specific diet, and it was found to be significant in the case of DHA (Figure 2C, p = 0.003), the DHA/DPA ratio (Figure 2D, p = 0.009), the total Ceramides (Figure 2E, p = 0.02), and the total dh-Ceramides (Figure 2F, p = 0.01). In summary, these findings indicate that a different dietary n-6/n-3 PUFA ratios and n-3 LC-PUFA content caused specific modifications in the brain lipidome of female APP/PS1 mice, which is partially dependent on the circulating levels of estradiol. To visualize the differential effects of these dietary regimes in female mice with different genotypes, we compared the relationship between the total n-6 FA content of the three diets used (SF, 10.63; DII, 16.55; DI, 28.84) and the mean cerebral cortex levels of some complex lipids in WT and APP/PS1 mice. Significant dose-response effects were seen in WT animals for Ceramides (p = 0.00001), dh- Ceramides (p = 0.000), Sphingomyelins (p = 0.00001), and dh-Sphingomyelins (p = 0.00001, Figure 3). By contrast, no relationship between these two variables was detected in APP/PS1 mice for any of the lipids analyzed. In addition, all the regression slopes were significantly different in the WT and APP/PS1 mice for Ceramides (p = 0.001), dh-Ceramides (p = 0.01), Sphingomyelins (p = 0.001), and dh-Sphingomyelins (p = 0.0009). How of the Interaction of the Different n-6/n-3 PUFA Ratios or n-3 LC-PUFA Content and the Gonadal Status Affects Hippocampal Accumulation of Amyloid Peptide Aβ1 –40 in Female APP/PS1 Mouse Changes in the dietary DHA intake reduce the amyloid FIGURE 1 | Effect of different dietary composition on cerebral cortex lipidome in intact female APP/PS1 mice. SF: Standard laboratory food (SF); DI: High burden in the brain of AD mouse models (Lim et al., 2005; n-6/n-3 ratio; DII: Low n-6/n-3 ratio. Data are represented in the units Green et al., 2007), and we previously showed that different indicated in the vertical axis as mean ± SEM of four mice per group. In this concentrations of n-6 and n-3 LC-PUFAs in the diet modify and other figures, the symbols indicate significant differences obtained from the brain levels of APP in WT mice (Herrera et al., 2018). the statistical analysis as described in methods section. (A) General ANOVA: Therefore, we analyzed the impact of both dietary composition p = 0.0004; (a): p = 0.0005 vs. SF; (B) General ANOVA: p = 0.001; (a): p = 0.0009 vs. SF or DII; (C) General ANOVA: p = 0.84; (D) General ANOVA: and gonadal status on brain amyloid accumulation in APP/PS1 p = 0.0001; (a): p = 0.0005, and (b) p = 0.02 vs. SF and DII; (E) General animals. The effects of the experimental diets on amyloid levels ANOVA: p = 0.0004; (a) p = 0.001 vs. SF; (F) General ANOVA: p = 0.0005; (a) were compared with that of SF and initially, no significant p = 0.05, and (b) p = 0.005 vs. SF; (G) General ANOVA: p = 0.18; (H) General differences in the levels the Aβ1–40 amyloid peptide were ANOVA: p = 0.005; (a) p = 0.02, and (b) p = 0.002 vs. SF; (I) General ANOVA: p = 0.03; (a) 0.04 vs. SF; (J) General ANOVA: p = 0.01; (a): p = 0.007 vs. SF. observed in the plasma with the three diets (Figure 4A, p = 0.34). However, there was less Aβ1–40 in the hippocampus Frontiers in Aging Neuroscience | www.frontiersin.org 6 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function FIGURE 2 | Differential effects of chronic estradiol administration on dietary-dependent cerebral cortex lipidome in ovariectomized female APP/PS1 mice. Clear bars: ovariectomized-placebo treated (OVX); Dark bars: ovariectomized-estradiol treated (OVX-E). Data are represented in the units indicated in the vertical axis as mean ± SEM of four mice per group. (A) General ANOVA: p = 0.05; (a) p = 0.02 vs. OVX; (B) General ANOVA: p = 0.00001; (a) p = 0.0001 vs. OVX; (C) General ANOVA: p = 0.0001; (D) General ANOVA: p = 0.0001; (a): p = 0.009 vs. OVX; (E) General ANOVA: p = 0.05; (a): p = 0.02 vs. OVX; (F) General ANOVA: p = 0.03; (a): 0.01 vs. OVX; (G) General ANOVA: p = 0.07; (a): p = 0.05 vs. OVX; (H) General ANOVA: p = 0.04; (a) p = 0.05 vs. OVX. of intact sham-operated females when animals were fed DII the levels of Aβ1–42among the different experimental groups rather than SF (Figure 4B, p = 0.017). While DI appeared (data not shown). to produce a similar effect, the difference between the DI and SF mice was not significant. Likewise, chronic estradiol administration did not significantly affect the hippocampal The Effects of Interaction Between amyloid accumulation in ovariectomized animals fed DI relative Different Dietary n-6/n-3 PUFA Ratios to those that received the placebo. By contrast, there was a and n-3 LC-PUFA Content With Gonadal significant reduction of hippocampal Aβ1–40 in OVX-E mice Status on the Expression of GFAP, fed DII relative to the OVX animals (Figure 4C, p = 0.01). Furthermore, the influence of this gonadal hormone on amyloid Signaling Proteins and Synaptic Markers levels depended on the diet, since the effect of estradiol in the Cerebral Cortex of Female in animals fed DI differed significantly from that observed APP/PS1 Mouse in the mice that received DII (Figure 4C: p = 0.01). In The expression of GFAP and of selected signaling proteins and contrast to Aβ1–40, no significant differences were detected in synaptic markers was compared between the cerebral cortex of Frontiers in Aging Neuroscience | www.frontiersin.org 7 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function FIGURE 3 | Linear regression between dietary content in n-6 FA and cerebral cortex levels of several complex lipids. Vertical axis represents the means of each compound in pmol/mg protein of four animals per group. Horizontal axis represents the progression of n-6 FA dietary content in gr/kg fresh weight (SF: 10.63; DII: 16.55; DI: 28.84). Significant dose-response effects are shown in intact WT mice for Ceramides (p = 0.00001), dh-Ceramides (p = 0.0004), Sphingomyelins (p = 0.0001), and dh-Sphingomyelins (p = 0.0001) but not in intact APP/PS1 mice. In addition, all slopes were significantly different between WT and APP/PS1 animals for Ceramides (p = 0.001), dh-Ceramides (p = 0.01), Sphingomyelins (p = 0.001), and dh-Sphingomyelins (p = 0.0009). FIGURE 4 | Effect of different dietary composition on plasma and hippocampal levels of Aβ1 - 40 in female APP/PS1 mice and its dependence on gonadal status. Data are represented in the units indicated in the vertical axis as mean ± SEM of five to six mice per group. (A) General ANOVA: p = 0.34; (B) General ANOVA: p = 0.048; (a) p = 0.017 vs. SF; (C) General ANOVA: p = 0.058; (a) p = 0.01 vs. OVX. animals fed DI and DII (animals maintained on SF were not or estradiol. However, the statistical analysis revealed a global included in these studies). When the dietary-induced changes enhancement in PI3K (p = 0.0001), Akt (p = 0.002), and GSK3 in the astrocyte marker GFAP were quantified in Western blots (p = 0.036) in the mice that received DII relative to those that (Figure 5A), more GFAP was detected in sham operated female received DI, irrespective of their gonadal status. In addition, APP/PS1 mice fed DII than in those fed DI (p = 0.002). This PI3K (p = 0.005), Akt (p = 0.00001), GSK3 (p < 0.00001), marked difference was not observed in ovariectomized animals, and GSK3-pSer (p = 0.005) were expressed more strongly irrespective of whether they receive the placebo or estradiol. In in intact sham-operated than in ovariectomized mice. This addition, GFAP expression in all the intact animals (SHAM) was global effect of ovariectomy was not prevented or reverted by found to be higher than in the ovariectomized mice, irrespective chronic estradiol administration since there were no significant of the diet (p = 0.0004) (Figure 5B). differences when OVX mice were compared with OVX-E mice. In terms of signaling proteins (Figure 6A), such as the As an exception to this general picture, DII significantly increased p85 subunit of PI3K, Akt, and GSK3 (total and serine the GSK3-pSer levels in OVX females relative to those that phosphorylated), their quantitative expression in Western blots received DI (p = 0.0005), an effect that was prevented by chronic was normalized to that of Actin (Figures 6B–E). Sham-operated estradiol treatment. animals fed DII expressed PI3K (p = 0.0001) and Akt (p = 0.003) We also analyzed the levels of some synaptic markers more strongly, yet these differences in the effect of diet were in the cerebral cortex, such as synapsin, p- synapsin, and not observed in ovariectomized mice treated with either placebo synaptophysin, proteins involved in presynaptic mechanisms Frontiers in Aging Neuroscience | www.frontiersin.org 8 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function Cederholm and Palmblad, 2010; Shinto et al., 2014) indicate that LC-PUFAs like DHA exert neuroprotective effects, preventing neurodegeneration and cognitive impairment. Indeed, diets enriched in DHA have been shown to reduce amyloidosis in AD mouse models (Lim et al., 2005; Green et al., 2007; Perez et al., 2010). In agreement with these studies, the results presented here indicate that a diet containing high levels of DHA and a low n-6/n-3 PUFA ratio reduces the amyloid Aβ1–40 in the hippocampus of female APP/PS1 mice when compared to diets with no traces of DHA and higher n-6/n-3 PUFA ratios. This effect we believe to be mainly due to increased 18:2n- 6 supplementation. The effect of the DHA enriched diet on amyloid Aβ1–40 levels was accompanied by modified expression of components of a neuroprotective signaling pathway implicated in the AD pathology, the PI3K/Akt/GSK3 cascade and of key synaptic proteins like synapsin and PSD95 (Gomez-Sintes et al., 2011; Reddy, 2013; Llorens-Martin et al., 2014; Beurel et al., 2015). Thus, a DHA enriched diet increased the expression of synapsin and PSD95, the synapsin/p-synapsin ratio, and that of PI3K in sham APP/PS1 mice. It is important to note that the levels of Aβ1–42, which is more prone to aggregate and form Aβ plaques, were not significantly modified by the diets. Thus, in the absence of an analysis of amyloid deposits it is not possible to determine the FIGURE 5 | Effect of different dietary composition on GFAP expression in the true implication of the diets in terms of Aβ plaque formation. cerebral cortex of female APP/PS1 mice under different gonadal conditions. (A) Western blots of GFAP and β-Actin. (B) Vertical axis represents relative However, PI3K is a neuroprotective kinase that promotes units of densitometric quantification as mean ± SEM of four mice per group, neuronal survival, memory, and cognition (Kitagishi et al., 2014; normalized with respect to the loading control, β-Actin. Clear bars correspond Arevalo et al., 2015; Matsuda et al., 2018). Similarly, PSD95 to high n-6/n-3 ratio (DI), and dark bars correspond to low n-6/n-3 ratio (DII). is an integral component of the postsynaptic density, which Horizontal axis indicates the different gonadal status as follows: SHAM, intact is involved in the development of new dendritic spines and sham-operated; OVX, ovariectomized, placebo-treated; OVX-E, ovariectomized, estradiol-treated. General ANOVA: p = 0.0004; (a) that fulfills important roles in synaptic plasticity (Kim and p = 0.0001. Sheng, 2004; Ehrlich et al., 2007; Berry and Nedivi, 2017). Synapsins are a family of presynaptic proteins that play critical roles in synaptic development, neurotransmitter release, and of neurotransmission, and PSD95 as a postsynaptic protein neural plasticity through phosphorylation-dependent processes target (Figure 7A). Sham operated mice fed DII presented (Chi et al., 2003; Menegon et al., 2006; Giachello et al., 2010). significantly higher levels of synapsin (p = 0.0008), synaptophysin Considering the roles of PI3K, PSD95, and synapsin, our results (p = 0.035), and PSD95 (p = 0.0004) and lower levels of suggest that diets enriched in n-3 LC-PUFA and with different p-synapsin (p = 0.0007) compared to mice fed DI (Figures 7B–E). ratios of specific n-6 and n-3 PUFAs modulate the expression of In the case of synapsin and p-synapsin, similar trends were key elements involved in neuroprotective signaling and synaptic observed in OVX mice although this trend did not reach transmission in AD mice, consistent with previous reports on significance. Nevertheless, a global effect of DII and DI was the effects of DHA-enriched diets on hippocampal neuronal observed for synapsin (p = 0.0003), p-synapsin (p = 0.001), development (Cao et al., 2009), synaptic function (Cao et al., synaptophysin (p = 0.05), and PSD95 (p = 0.003) irrespective 2009), synaptic membrane proteins (Calon et al., 2004; Cansev of gonadal status. Moreover, as for the signaling proteins, et al., 2008; Sidhu et al., 2016; Herrera et al., 2018), and ovariectomy was associated with reduced synapsin (p = 0.0001), the restoration of neural plasticity and cognition after brain p-synapsin (p = 0.001), synaptophysin (p < 0.00001), and PSD95 trauma (Wu et al., 2011). Our present findings indicate that (p < 0.00001) expression, independent of the diet. As mentioned two diets differing in their specific ARA, EPA, DPA, and DHA above, this effect was not affected by chronic estradiol treatment. content, and in the n-6/n-3 PUFA ratio induce divergent changes in the lipidome of AD mouse brains. Compared to SF, DI increases the brain ARA and DPA levels while significantly DISCUSSION reducing DHA levels, while DII caused a decrease in DPA and a significant increase in the DHA/DPA ratio. The different Previous experimental (Calon et al., 2004; Lim et al., 2005; changes in ARA and DHA induced by the two diets are Green et al., 2007; Cao et al., 2009; Hooijmans et al., 2009) consistent with the differences in their composition because and epidemiological data (Kalmijn et al., 1997; Barberger- in rodents these lipids can be obtained either from the diet Gateau et al., 2002; Engelhart et al., 2002; Morris et al., 2005; or they can be synthesized in the liver from linoleic acid Frontiers in Aging Neuroscience | www.frontiersin.org 9 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function FIGURE 6 | Effect of different dietary composition on the expression of signaling proteins in the cerebral cortex of female APP/PS1 mice under different gonadal conditions. (A) Western blots for PI3K, AKT, GSK3, GSK3-pSer (p-GSK3), and β-Actin. (B–E) Vertical axis represents relative units of densitometric quantification as mean ± SEM of four mice per group, normalized with respect to the loading control, β-Actin. Clear bars correspond to high n-6/n-3 ratio (DI), and dark bars correspond to low n-6/n-3 ratio (DII). Horizontal axis indicates the different gonadal status as follows: SHAM, intact sham-operated; OVX, ovariectomized, placebo-treated; OVX-E, ovariectomized, estradiol-treated. (B) General ANOVA: p = 0.000001; (a) p = 0.00001; Global effect of DII vs. DI: p = 0.0003. (C) General ANOVA: p = 0.02; (a) p = 0.0004; Global effect of DII vs. DI: p = 0.01. (D) General ANOVA: p = 0.02; (a) p = 0.03; (b) p = 0.008; Global effect of DII vs. DI: p = 0.002. (E) General ANOVA: p = 0.0004; (a) p = 0.0005. (18:2n-6) or α-linolenic acid (18:3n-3), respectively (Sprecher, genotypes in terms of total FA, the relative levels of ARA, 2000; Williard et al., 2001; Gregory et al., 2011; Domenichiello DPA, DHA, and the ratio of DHA/C22:5n-6. However, certain et al., 2014). In humans, although DHA brain metabolism quantitative differences were observed. For instance, while the remains to be completely understood, the situation seems to be full repertoire of Ceramides, dh-Ceramides, Sphingomyelins, and similar in that dietary intake of its precursors may be sufficient dh-Sphingomyelins appeared to be elevated by both DI and DII to supply the adult brain (Barceló-Coblijn and Murphy, 2009; diets in APP/PS1 mice, the effect of DII was significantly stronger Domenichiello et al., 2015). than that of DI, in contrast to the effect observed in WT animals The dietary effects on the lipid composition of the female (Herrera et al., 2018). Equivalent lipidome remodeling has been APP/PS1 mouse brain detected here are in some aspects similar reported in the hippocampus of female animals of the same AD to those observed previously in WT mice maintained on the model fed similar diets (Diaz et al., 2016), with WT and AD mice same diets (Herrera et al., 2018). To compare the results directly also suffering similar changes in the hippocampal lipid profile and with those of our previous study on WT animals we used the certain differences depending on the genotype. Here, the main same experimental diets (DI, DII) and SF (Herrera et al., 2018). quantitative differences in the cerebral cortex between APP/PS1 The diets have a similar effect on the brain lipidome in both mice and WT mice were found in complex lipids, suggesting that Frontiers in Aging Neuroscience | www.frontiersin.org 10 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function FIGURE 7 | Effect of different dietary composition on the expression of synaptic proteins in the cerebral cortex of female APP/PS1 mice under different reproductive conditions. (A) Western blots for synapsin, p-synapsin, synaptophysin, PSD95, and β-Actin. (B–E) Vertical axis represents relative units of densitometric quantification as mean ± SEM of four mice per group, normalized with respect to the loading control, β-Actin. Clear bars correspond to high n-6/n-3 ratio (DI), and dark bars correspond to low n-6/n-3 ratio (DII). Horizontal axis indicates the different gonadal status as follows: SHAM, intact sham-operated; OVX, ovariectomized, placebo-treated; OVX-E, ovariectomized, estradiol-treated. (B) General ANOVA: p = 0.0005; (a) p = 0.01; (b) p = 0.001. (C) General ANOVA: p = 0.0006; (a) p = 0.003; (b) p = 0.01; (c) p = 0.0005; Global effect of DII vs. DI: p < 0.0000001. (D) General ANOVA: p = 0.0005; (a) p = 0.0006; (b) p = 0.003; (c) p = 0.009; Global effect of DII vs. DI: p < 0.0000001. (E) General ANOVA: p = 0.008; (a) p = 0.007. dietary composition may significantly affect lipid metabolism in on the protein profile of WT female mice (Herrera et al., 2018). the AD mouse brain. Therefore, the main aim of the present study was to investigate Neuroprotective effects of ovarian hormones, in particular the consequences of such an interaction between ovarian estradiol, are well characterized, and they include the function and DHA-enriched diets in the AD mouse brain. downregulation of amyloidosis (Pike et al., 2009), the modulation Indeed, an interesting conclusion drawn from our results is that of PI3K signaling (Marin et al., 2005; Arevalo et al., 2015), and the dietary effects on the cerebral cortex lipidome in APP/PS1 the regulation of synaptic proteins like PSD95 (Luine et al., mice differ depending on the levels of circulating estradiol. 2018) or synapsin (Rebas et al., 2005). All these parameters Thus, in ovariectomized AD animals fed DI, estradiol increased are modified by a DHA-enriched diet, as suggested in the the brain levels of total FAs, DPA, Ceramides, dh-Ceramides, present study, although the modification induced by the diet on Sphingomyelins, and dh-Sphingomyelins. In addition, estradiol synaptic proteins were not very large. Considering the hormonal increased the DHA/DPA ratio in the brain of ovariectomized decline associated with the menopause, an important unexplored AD mice maintained on DII. These results extend those of our question is whether ovarian function affects the neuroprotective previous studies showing that ovariectomy profoundly alters the effects of DHA, as suggested previously when a DHA enriched dietary effects on the lipidome in the cerebral cortex of WT mice diet was seen to diminish Aβ plaque deposition in female but (Herrera et al., 2018) and in the hippocampus of APP/PS1 mice not male AD mice (Perez et al., 2010). Moreover, ovariectomy (Diaz et al., 2016). However, one limitation of our study is that prevents the effects of this type of diet on the brain lipidome and the lipidome of intact non-ovariectomized APP/PS1 animals was Frontiers in Aging Neuroscience | www.frontiersin.org 11 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function not analyzed. Nevertheless, these findings together indicate that ETHICS STATEMENT the effect of diet on the brain lipidome is influenced by ovarian function and circulating estradiol levels in WT and AD mice. The The animal study was reviewed and approved by the University influence of ovarian hormones on the effect of the diet on the of La Laguna Animal Care and Use Committee. mouse brain lipidome is consistent with human studies showing that estrogen stimulates the conversion of n-3 and n-6 precursors into their long chain metabolites (Decsi and Kennedy, 2011). AUTHOR CONTRIBUTIONS Therefore, the potential synergies between diet and gonadal function should be carefully considered when designing dietary JH performed the animal experiments and was in charge of therapeutic approaches for neurological diseases. most experimental manipulations, hormone treatments, and As observed for the changes in the brain lipidome, the dietary collection of brain samples. JH and LO-G prepared the brain effect on amyloid Aβ1–40 levels in APP/PS1 animals was also samples and performed the western blotting analysis. AM and dependent on circulating estradiol, indicating that there is an GH participated in the experimental designs, animal treatments, interaction between the protective effect of estradiol and the and collection of samples. GF and JC performed the lipidomic intake of a DHA-enriched diet with a low n-6/n-3 PUFA ratio in analysis of brain samples. NA and CR analyzed the lipid an AD animal model. The effect of the DHA-enriched diet on the composition of the experimental diets. CR collaborated in the expression of some components of the PI3K/Akt/GSK3 pathway interpretation of results of brain lipidome. LP-V performed the was also dependent on the hormonal status of AD mice. Thus, the statistical analysis of quantitative results. RA conceived and effect of the DHA-enriched diet on the expression of Akt was lost designed the experiments and contributed together with LP-V to in ovariectomized AD animals, while this diet only increased the the design and interpretation of specific statistical analysis. RA, GSK3 levels in ovariectomized animals. Furthermore, estradiol LG-S, and FW were in charge of the final interpretation of results prevented the effect of the DHA-enriched diet on Akt and serine and wrote the manuscript. phosphorylated GSK3 in ovariectomized mice. Therefore, our present findings suggest that this hormone interacts with the DHA-enriched and a low n-6/n-3 PUFA ratio diet to regulate key components of this signaling pathway in the brain of AD mice. FUNDING Similar conclusions can be drawn regarding the interaction of RA was supported by SAF2007-66148-C02-01, FP7-REGPOT- diet and estradiol on the regulation of synaptic proteins. Thus, 2012-31637-IMBRAIN grants and Fundación Canaria para el the effect of the DHA-enriched diet on PSD95 and synapsin levels Avance de la Medicina y la Biotecnología. LG-S was supported in the cerebral cortex of AD mice was prevented by ovariectomy by the Agencia Estatal de Investigación (BFU2017-82754- or by estradiol administration, respectively. R), Centro de Investigación Biomédica en Red Fragilidad y The interaction of these two factors, diet and gonadal status, Envejecimiento Saludable (CIBERFES), and Fondos FEDER. FW was also observed when analyzing the cerebral cortex levels was supported by grants from the MICINN (SAF2015-70368-R), of GFAP as an index of astroglial activation. Together, these CAM (B2017/BMD-3700), Centro de Investigación Biomédica results suggest that a diet enriched in DHA and a low n- en Red sobre Enfermedades Neurodegenerativas (CIBERNED), 6/n-3 ratio did not safeguard the beneficial effects on amyloid and an institutional grant from the Fundación Ramón Areces to peptide accumulation, neuroprotective signaling, and synaptic CBMSO and Fondos FEDER. function per se, independently of circulating estradiol. Only in combination with estradiol does a diet enriched in DHA and with a low n-6/n-3 ratio profoundly change Aβ metabolism in this AD model. However, in terms of the implications of our findings ACKNOWLEDGMENTS for AD, one limitation of the present study is that the APP/PS1 mouse model does not reproduce all the pathological hallmarks The authors gratefully acknowledge Mario Diaz (Universidad of the disease, such as tau pathology. Hence, further studies in de La Laguna) and Emilio Martinez de la Victoria (Instituto other mouse models of AD will be necessary. Nevertheless, our de Nutrición y Tecnología de los Alimentos, Universidad de study does support the idea that the design of dietary and/or Granada) for their advice in the design of experimental diets, and pharmacological interventions for human neurodegenerative Maria Rosa Arnaiz for her help in the management of mice in the diseases should take into account the hormonal and metabolic University of La Laguna animal facility. status that may affect the therapeutic response. SUPPLEMENTARY MATERIAL DATA AVAILABILITY STATEMENT The Supplementary Material for this article can be found The datasets generated for this study are available on request to online at: https://www.frontiersin.org/articles/10.3389/fnagi. the corresponding author. 2019.00346/full#supplementary-material Frontiers in Aging Neuroscience | www.frontiersin.org 12 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function REFERENCES Diaz, M., Fabelo, N., Casañas-Sanchez, V., Marin, R., Gomez, T., Quinto-Alemany, D., et al. (2016). Hippocampal lipid homeostasis in APP/PS1 mice is modulated Arevalo, M. A., Azcoitia, I., Gonzalez-Burgos, I., and Garcia-Segura, L. M. (2015). by a complex interplay between dietary DHA and estrogens: relevance for Signaling mechanisms mediating the regulation of synaptic plasticity and Alzheimer’s disease. J. Alzheimer’s Dis. 49, 459–481. doi: 10.3233/JAD-150470 memory by estradiol. Horm. Behav. 74, 19–27. doi: 10.1016/j.yhbeh.2015.04016 Domenichiello, A. F., Chen, C. T., Trepanier, M. O., Stauro, P. M., and Bazinet, Armitage, P., Berry, G., and Mathews, J. N. S. (2004). Statistical Methods in Medical R. P. (2014). Whole body synthesis rates of DHA from alpha-linolenic acid are Research. London: Wiley-Blackwell. greater than brain DHA accretion and uptake rates in adult rats. J. Lipid Res. 55, Baglietto-Vargas, D., Shi, J., Yaeger, D. M., Ager, R., and LaFerla, F. M. (2016). 62–74. doi: 10.1194/jlr.M042275 Diabetes and Alzheimer’s disease crosstalk. Neurosci. Biobehav. Rev. 64, Domenichiello, A. F., Kitson, A. P., and Bazinet, R. P. (2015). Is docosahexaenoic 272–287. doi: 10.1016/j.neubiorev.2016.03.005 acid synthesis alpha-linolenic acid sufficient to supply the adult brain? Progr. Bannenberg, G., and Serhan, C. N. (2010). Specialized pro-resolving lipid Lipid Res. 59, 54–66. doi: 10.1016/j.plipres.2015.04.002 mediators in the inflammatory response: an update. Biochim. Biophys. Acta Duyckaerts, C. (2003). [What is dementia? 3. The neuropathologist point of view]. 1801, 1260–1273. doi: 10.1016/j.bbalip.2010.08002 Psychol. Neuropsychiatr. Vieill. 1, 71–77. Barberger-Gateau, P., Letenneur, L., Deschamps, V., Peres, K., Dartigues, J. F., and Ehrlich, I., Klein, M., Rumpel, S., and Malinow, R. (2007). PSD-95 is required Renaud, S. (2002). Fish, meat, and risk of dementia: cohort study. BMJ 325, for activity-driven synapse stabilization. Proc. Natl. Acad. Sci. U.S.A. 104, 932–933. doi: 10.1136/bmj.325.7370.932 4176–4181. doi: 10.1073/pnas.0609307104 Barceló-Coblijn, G., and Murphy, E. J. (2009). Alpha-linolenic acid and its Engelhart, M. J., Geerlings, M. I., Ruitenberg, A., van Swieten, J. C., Hofman, conversion to long-chain n-3 fatty acids: benefits for human health and a A., Witteman, J. C., et al. (2002). Dietary intake of antioxidants and risk of role in maintaining tissue n-3 fatty acid levels. Progr. Lipid Res. 48, 355–374. Alzheimer disease. JAMA 287, 3223–3229. doi: 10.1001/jama.287.24.3223 doi: 10.1016/j.plipres.2009.07.002 Fabelo, N., Martin, V., Marin, R., Moreno, D., Ferrer, I., and Diaz, M. (2014). Barha, C. K., and Galea, L. A. (2010). Influence of different estrogens on Altered lipid composition in cortical lipid rafts occurs at early stages of sporadic neuroplasticity and cognition in the hippocampus. Biochim. Biophys. Acta 1800, Alzheimer’s disease and facilitates APP/BACE1 interactions. Neurobiol. Aging 1056–1067. doi: 10.1016/j.bbagen.2010.01.006 35, 1801–1812. doi: 10.1016/j.neurobiolagiung.2014.02.005 Berry, K. P., and Nedivi, E. (2017). Spine dynamics: are they all the same? Neuron Ferreira, L. S. S., Fernandes, C. S., Vieira, M. N. N., and De Felice, F. G. (2018). 27, 43–55. doi: 10.1016/j.neuron.2017.08.008 Insulin resistance in Alzheimer’s disease. Front. Neurocsi. 13:830. doi: 10.3389/ Beurel, E., Grieco, S. F., and Jope, R. S. (2015). Glycogen synthase kinase-3 (GSK3): fnins.2018.00830 regulation, actions, and diseases. Pharmacol. Ther. 148, 114–131. doi: 10.1016/ Folch, J., Lees, M., and Sloane Stanley, S. S. (1957). A simple method for the j.pharmthera.2014.11.016 isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226, Boccardi, V., Murasecco, I., and Mecocci, P. (2019). Diabetes drugs in the fight 497–509. against Alzheimer’s disease. Ageing Res. Rev. 54:100936. doi: 10.1016/j.arr.2019. Freund-Levi, Y., Eriksdotter-Jonhagen, M., Cederholm, T., Basun, H., Faxen- 100936 Irving, G., Garlind, A., et al. (2006). Omega-3 fatty acid treatment in Calon, F., Lim, G. P., Yang, F., Morihara, T., Teter, B., Ubeda, O., et al. (2004). 174 patients with mild to moderate Alzheimer disease: OmegAD study: a Docosahexaenoic acid protects from dendritic pathology in an Alzheimer’s randomized double-blind trial. Arch. Neurol. 63, 1402–1408. doi: 10.1001/ disease mouse model. Neuron 43, 633–645. doi: 10.1016/j.neuron.2004.08.013 archneur.63.10.1402 Cansev, M., Ulus, I. H., Wang, L., Maher, T. J., and Wurtman, R. J. (2008). Giachello, C. N., Fiumara, F., Giacomini, C., Corradi, A., Milanese, C., Ghirardi, Restorative effects of uridine plus docosahexaenoic acid in a rat model of M., et al. (2010). MAPK/Erk-dependent phosphorylation of synapsin mediates Parkinson’s disease. Neurosci. Res. 62, 206–209. doi: 10.1016/j.neures.2008. formation of functional synapses and short-term homosynaptic plasticity. 07.005 J. Cell. Sci. 123, 881–893. doi: 10.1242/jcs.056846 Cao, D., Kevala, K., Kim, J., Moon, H. S., Jun, S. B., Lovinger, D., et al. Gibson, G. E., and Huang, H. M. (2005). Oxidative stress in Alzheimer’s disease. (2009). Docosahexaenoic acid promotes hippocampal neuronal development Neurobiol. Aging 26, 575–578. doi: 10.1016/j.neurobiolaging.2004.07.017 and synaptic function. J. Neurochem. 111, 510–521. doi: 10.1111/j.1471-4159. Gomez-Sintes, R., Hernandez, F., Lucas, J. J., and Avila, J. (2011). GSK-3 mouse 2009.06335.x models to study neuronal apoptosis and neurodegeneration. Front. Mol. Catalan, J., Moriguchi, T., Slotnick, B., Murthy, M., Greiner, R. S., and Salem, Neurosci. 4:45. doi: 10.3389/fnmol.2011.00045 N. (2002). Cognitive deficits in docosahexaenoic acid-deficient rats. Behav. Green, K. N., Martinez-Coria, H., Khashwji, H., Hall, E. B., Yurko-Mauro, K. A., Neurosci. 116, 1022–1031. doi: 10.1016/S0031-9384(00)00437-6 Ellis, L., et al. (2007). Dietary docosahexaenoic acid and docosapentaenoic Cederholm, T., and Palmblad, J. (2010). Are omega-3 fatty acids options for acid ameliorate amyloid-beta and tau pathology via a mechanism involving prevention and treatment of cognitive decline and dementia? Curr. Opin. Clin. presenilin 1 levels. J. Neurosci. 27, 4385–4395. doi: 10.1523/JNEUROSCI.0055- Nutr. Metab. Care 13, 150–155. doi: 10.1097/MCO.0b013e328335c40b 07.2007 Chi, P., Greengard, P., and Ryan, T. A. (2003). Synaptic vesicle mobilization Gregory, M. K., Gibson, R. A., Cook-Johnson, R. J., Cleland, L. G., and James, M. J. is regulated by distinct synapsin I phosphorylation pathways at different (2011). Elongase reactions as control points in long-chain polyunsaturated fatty frequencies. Neuron 38, 69–78. doi: 10.1016/s0896-6273(03)00151-x acid synthesis. PLoS One 6:e29662. doi: 10.1371/journal.pone.0029662 Cingolani, F., Casasampere, M., Sanllehi, P., Casas, J., Bujons, J., and Fabrias, G. Guerra, B., Diaz, M., Alonso, R., and Marin, R. (2004). Plasma membrane (2014). Inhibition of dihydroceramide desaturase activity by the sphingosine oestrogen receptor mediates neuroprotection against beta-amyloid toxicity kinase inhibitor SKI II. J. Lipid Res. 55, 1711–1720. doi: 10.1194/jlr.M049759 through activation of Raf-1/MEK/ERK cascade in septal-derived cholinergic Dar, T. A., Sheikh, I. A., Ganie, S. A., Ali, R., Singh, L. R., Gan, S. H., et al. SN56 cells. J. Neurochem. 91, 99–109. doi: 10.1111/j.1471-4159.2004.02695.x (2014). Molecular linkages between diabetes and Alzheimer’s disease: current Hardy, J., and Selkoe, D. J. (2002). The amyloid hypothesis of Alzheimer’s disease: scenario and future prospects. CNS Neurol. Disord. Drug Targets 13, 290–298. progress and problems on the road to therapeutics. Science 297, 353–356. doi: doi: 10.2174/18715273113126660135 10.1126/science.1072994 Das, U., Scott, D. A., Ganguly, A., Koo, E. H., Tang, Y., and Roy, S. (2013). Herrera, J. L., Ordoñez-Gutierrez, L., Fabrias, G., Casas, J., Morales, A., Hernandez, Activity-induced convergence of APP and BACE-1 in acidic microdomains via G., et al. (2018). Ovarian function modulates the effects of long-chain an endocytosis-dependent pathway. Neuron 79, 447–460. doi: 10.1016/j.neuron. polyunsaturated fatty acids on the mouse cerebral cortex. Front. Cell. Neurosci. 2013.05.035 12:103. doi: 10.3389/fncel.2018.00103 Das, U., Wang, L., Ganguly, A., Saikia, J. M., Wagner, S. L., Koo, E. H., et al. Hicks, D. A., Nalivaieva, N. N., and Turner, A. J. (2012). Lipid rafts and Alzheimer’s (2016). Visualizing APP and BACE-1 approximation in neurons yields insight disease: protein-lipid interactions and perturbation of signaling. Front. Physiol. into the amyloidogenic pathway. Nat. Neurosci. 19, 55–64. doi: 10.1038/nn. 3:189. doi: 10.3389/fphys.2012.00189 4188 Hooijmans, C. R., Ven der Zee, C. E., Dederen, P. J., Brower, K. M., Reijmer, Decsi, T., and Kennedy, K. (2011). Sex-specific differences in essential fatty acid Y. D., van Groen, T., et al. (2009). DHA and cholesterol containing diets metabolism. Am. J. Clin. Nutr. 94, 1914S–1919S. doi: 10.3945/ajcn.110.000893 influence Alzheimer-like pathology, cognition and cerebral vasculature in Frontiers in Aging Neuroscience | www.frontiersin.org 13 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function APPswe/PS1dE9 mice. Neurobiol. Dis. 33, 482–498. doi: 10.1016/j.nbd.2008. aromatase gene with Alzheimer’s disease and its interaction with IL10 in the 12.002 Epistasis project. Eur. J. Hum. Genet. 22, 216–220. doi: 10.1038/ejhg.2013.116 Huang, R., and Poduslo, S. E. (2006). CYP19 haplotypes increase risk for Menegon, A., Bonanomi, D., Albertinazzi, C., Lotti, F., Ferrari, G., Kao, H. T., Alzheimer’s disease. J. Med. Genet. 43:e42. doi: 10.1136/jmg.2005.039461 et al. (2006). Protein kinase A-mediated synapsin I phosphorylation is a central. Iivonen, S., Corder, E., Lehtovirta, M., Helisalmi, S., Mannermaa, A., Vepsalainen, Modulator of Ca2+-dependent synaptic activity. J. Neurosci. 26, 11670–11681. S., et al. (2004). Polymorphisms in the CYP19 gene confer increased risk for doi: 10.1523/JNEUROSCI.3321-06.2006 Alzheimer disease. Neurology 62, 1170–1176. doi: 10.1212/01WNL.0000118208. Morris, M. C., Evans, D. A., Tangney, C. C., Bienias, J. L., and Wilson, R. S. (2005). 16939.60 Fish consumption and cognitive decline with age in a large community study. Ikemoto, A., Ohishi, M., Sato, Y., Hata, N., Misawa, Y., Fujii, Y., et al. (2001). Arch. Neurol. 62, 1849–1853. doi: 10.1001/archneur.62.12.noc.5061 Reversibility of n-3 fatty acid deficiency-induced alterations of learning Mullan, M., Crawford, F., Axelman, K., Houlden, H., Lilius, L., Winblad, B., et al. behavior in the rat: level of n-6 fatty acids as another critical factor. J. Lipid (1992). A pathogenic mutation for probable Alzheimer’s disease in the APP Res. 42, 1655–1663. gene at the N-terminus of beta-amyloid. Nat. Genet. 1, 345–347. doi: 10.1038/ Janicki, S. C., Park, N., Cheng, R., Schupf, N., Clark, L. N., and Lee, J. H. (2013). ng089-345 Aromatase variants modify risk for Alzheimer’s disease in a multiethnic female Orr, S. K., Palumbo, S., Bosetti, F., Mount, H. T., Kang, J. X., Greenwood, C. E., et al. cohort. Dement. Geriatr. Cogn. Disord. 35, 340–346. doi: 10.1159/000343074 (2013). Unesterified docosahexaenoic acid is protective in neuroinflammation. Jankowsky, J. L., Slunt, H. H., Ratovitski, T., Jenkins, N. A., Copeland, N. G., and J. Neurochem. 127, 378–393. doi: 10.1111/jnc.12392 Borchelt, D. R. (2001). Co-expression of multiple transgenes in mouse CNS: Overk, C. R., Lu, P. Y., Wang, Y. T., Choi, J., Shaw, J. W., Thatcher, G. R., et al. a comparison of strategies. Biomol. Eng. 17, 157–165. doi: 10.1016/s1389- (2012). Effects of aromatase inhibition versus gonadectomy on hippocampal 0344(01)00067-3 complex amyloid pathology in triple transgenic mice. Neurobiol. Dis. 45, Kalmijn, S., Launer, L. J., Ott, A., Witteman, J. C., Hofman, A., and Breteler, M. M. 479–487. doi: 10.1016/j.nbd.2011.08.035 (1997). Dietary fat intake and the risk of incident dementia in the Rotterdam Park, S. H., Kim, J. H., Choi, K. H., Jang, Y. J., Bae, S. S., Choi, B. T., et al. (2013). Study. Ann. Neurol. 42, 776–782. doi: 10.1002/ana.410420514 Hypercholesterolemia accelerates amyloid beta-induced cognitive deficits. Int. Kandimalla, R., Thirumala, V., and Reddy, P. H. (2017). Is Alzheimer’s disease a J. Mol. Med. 31, 577–582. doi: 10.3892/ijmm.2013.1233 Type 3 Diabetes? A critical appraisal. Biochim. Biophys. Acta Mol. Basis Dis. Perez, S. E., Berg, B. M., Moore, K. A., He, B., Counts, S. E., Fritz, J. J., et al. (2010). 1863, 1078–1089. doi: 10.1016/j.bbadis.2016.08.018 DHA diet reduces AD pathology in young APPswe/PS1(E9 transgenic mice: Kim, E., and Sheng, M. (2004). PDZ domain proteins of synapses. Nat. Rev. possible gender effects. J. Neurosci. Res. 88, 1026–1040. doi: 10.1002/jnr.22266 Neurosci. 5, 771–781. doi: 10.1038/nm1517 Pike, C. J., Carroll, J. C., Rosario, E. R., and Barron, A. M. (2009). Protective Kitagishi, Y., Nakanishi, A., Ogura, Y., and Matsuda, S. (2014). Dietary regulation actions of sex steroid hormones in Alzheimer’s disease. Front. Neuroendocrinol. of PI3K/AKT/GSK-3(pathway in Alzheimer’s disease. Alzheimers Res. Ther. 30:239–258. doi: 10.106/j.yfrn.2009.04.015 6:35. doi: 10.1186/alzrt265 Pugazhenthi, S., Qin, L., and Reddy, P. H. (2017). Common neurodegenerative Levy-Lahad, E., Wasco, W., Poorkaj, P., Romano, D. M., Oshima, J., Pettingell, pathways in obesity, diabetes, and Alzheimer’s disease. Biochim. Biophys. Acta W. H., et al. (1995). Candidate gene for the chromosome 1 familial Alzheimer’s Mol. Basis Dis. 1863, 1037–1045. doi: 10.1016/j.bbadis.2016.04.017 disease locus. Science 269, 973–977. doi: 10.1126/science.7638622 Rebas, E., Lachowicz, L., and Lachovicz, A. (2005). Estradiol modulates the Lim, G. P., Calon, F., Marihara, T., Yang, F., Teter, B., Ubeda, O., et al. (2005). synapsins phosphorylation by various protein kinases in the rat brain under A diet enriched with the omega-3 fatty acid docosahexaenoic acid reduces in vitro and in vivo conditions. J. Physiol. Pharmacol. 56, 39–48. amyloid burden in an aged Alzheimer mouse model. J. Neurosci. 25, 3032–3040. Reddy, P. H. (2013). Amyloid beta-induced glycogen synthase kinase 3β doi: 10.1523/JNEUROSCI.4225-04.2005 phosphorylated VDAC in Alzheimer’s disease: implications for synaptic Llorens-Martin, M., Jurado, J., and Avila, J. (2014). GSK-3ß, a pivotal kinase in dysfunction and neuronal damage. Biochem. Biophys. Acta 1832, 1–23. doi: Alzheimer disease. Front. Mol. Neurosci. 7:46. doi: 10.3389/fnmol.2014.00046 10.1016/j.bbadis.2013.06.012 Luine, V., Serrano, P., and Frankfurt, M. (2018). Rapid effect on memory Ricciarelli, R., Canepa, E., Marengo, B., Marinari, U. M., Poli, G., Pronzato, M. A., consolidation and spine morphology by estradiol in female and male rodents. et al. (2012). Cholesterol and Alzheimer’s disease: a still poorly understood Horm. Behav. 104, 111–118. doi: 10.1016/j.yhbeh.2018.04.007 correlation. IUBMB Life 64, 931–935. doi: 10.1002/iub.1091 Manly, J. J., Merchant, C. A., Jacobs, D. M., Small, S. A., Bell, K., Ferin, M., Salas, I. H., and De Strooper, B. (2019). Diabetes and Alzheimer’s disease: a link not et al. (2000). Endogenous estrogen levels and Alzheimer’s disease among as simple as it seems. Neurochem. Res. 44, 1271–1278. doi: 10.1007/s11064-018- postmenopausal women. Neurology 54, 833–837. doi: 10.1212/WNL.54.4.833 2690-9 Marin, R., Guerra, B., Alonso, R., Ramirez, C. M., and Diaz, M. (2005). Estrogen Scarmeas, N., Stern, Y., Tang, M. X., Mayeux, R., and Luchsinger, J. A. (2006). activates classical and alternative mechanisms to orchestrate neuroprotection. Mediterranean diet and risk for Alzheimer’s disease. Ann. Neurol. 59, 912–921. Curr. Neurovasc. Res. 2, 287–301. doi: 10.2174/156720205774322629 doi: 10.1002/ana.20854 Marin, R., Guerra, B., Morales, A., Diaz, M., and Alonso, R. (2003). An Selkoe, D. J. (2002). Alzheimer’s disease is a synaptic failure. Science 298, 789–791. estrogen membrane receptor participates in the neuroprotective action doi: 10.1126/science.1074069 of estradiol against amyloid β-peptide1-40-induced toxicity in SN56 Serhan, C. N. (2014). Pro-resolving lipid mediators are leads for resolution neurons. J. Neurochem. 85, 1180–1189. doi: 10.1046/j.1471-4159.2003. physiology. Nature 510, 92–101. doi: 10.1038/nature13479 01767.x Serhan, C. N., and Chiang, N. (2013). Resolution phase lipid mediators of Matsuda, S., Kakagawa, Y., Tsuji, A., Kitagishi, Y., Nakanishi, A., and Murai, T. inflammation: agonists of resolution. Curr. Opin. Pharmacol. 13, 632–640. doi: (2018). Implications of PI3K/AKT/PTEN signaling on superoxide dismutases 10.1016/j.coph.2013.05.012 expression and in the pathogenesis of Alzheimer’s disease. Diseases 6:28. doi: Sherrington, R., Froelich, S., Sorbi, S., Campion, D., Chi, H., Rogaeva, E. A., et al. 10.3390/diseases6020028 (1996). Alzheimer’s disease associated with mutations in presenilin 2 is rare Mazza, M., Pomponi, M., Janiri, L., Bria, P., and Mazza, S. (2007). Omega-3 fatty and variably penetrant. Hum. Mol. Genet. 5, 985–988. doi: 10.1093/hmg/5. acids and antioxidants in neurological and psychiatric diseases: an overview. 7.985 Prog. Neuropsychopharmacol. Biol. Psychiatry 31, 12–26. doi: 10.1016/j.pnpbp. Shinto, L., Quinn, J., Montine, T., Dodge, H. H., Woodward, W., Baldauf-Wagner, 2006.07.010 S., et al. (2014). A randomized placebo-controlled pilot trial of omega-3 fatty McCullough, L. D., Blizzard, K., Simpson, E. R., Oz, O. K., and Hurn, P. D. (2003). acids and alpha lipoic acid in Alzheimer’s disease. J. Alzheimers Dis. 38, 111–120. Aromatase cytochrome P450 and extragonadal estrogen play a role in ischemic doi: 10.3233/JAD-130722 neuroprotection. J. Neurosci. 23, 8701–8705. doi: 10.1523/jneurosci.23-25- Sidhu, V. K., Huang, B. X., Desai, A., Kevala, K., and Kim, H. Y. (2016). Role 08701.2003 of DHA in aging-related change in mouse brain synaptic plasma membrane Medway, C., Combarros, O., Cortina-Borja, M., Butler, H. T., Ibrahim-Verbaas, proteome. Neurobiol. Aging 41, 73–85. doi: 10.1016/j.neurobiolaging.2016. C. A., de Bruijn, R. F., et al. (2014). The sex-specific associations of the 02.007 Frontiers in Aging Neuroscience | www.frontiersin.org 14 December 2019 | Volume 11 | Article 346 Herrera et al. DHA, Alzheimer and Ovarian Function Sprecher, H. (2000). Metabolism of highly unsaturated n-3 and n-6 fatty Wu, A., Ying, Z., and Gomez-Pinilla, F. (2011). The salutary effects of DHA dietary acids. Biochem. Biophys. Acta 1486, 219–231. doi: 10.1016/s1388-1981(00) supplementation on cognition, neuroplasticity, and membrane homeostasis 00077-9 after brain trauma. J. Neurotrauma 28, 2113–2122. doi: 10.1089/neu.2011.1872 Talbot, K., Wang, H. Y., Kazi, H., Han, L. Y., Bakshi, K. P., Stucky, A., et al. Xu, H., Gouras, G. K., Greenfield, J. P., Vincent, B., Naslund, J., Mazzarelli, L., (2012). Demonstrated brain insulin resistance in Alzheimer’s disease patients et al. (1998). Estrogen reduces neuronal generation of Alzheimer beta-amyloid is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. peptides. Nat. Med. 4, 447–451. doi: 10.1038/nm0498-447 J. Clin. Invest. 122, 1316–1338. doi: 10.1172/JCI59903 Trinchese, F., Liu, S., Battaglia, F., Walter, S., Mathews, P. M., and Arancio, Conflict of Interest: The authors declare that the research was conducted in the O. (2004). Progressive age-related development of Alzheimer-like absence of any commercial or financial relationships that could be construed as a pathology in APP/PS1 mice. Ann. Neurol. 55, 801–814. doi: 10.1002/ana. potential conflict of interest. 20101 Viana da Silva, S., Haberl, M. G., Zhang, P., Bethge, P., Lemos, C., Gonçalves, N., Copyright © 2019 Herrera, Ordoñez-Gutierrez, Fabrias, Casas, Morales, Hernandez, et al. (2016). Early synaptic deficits in the APP/PS1 mouse model of Alzheimer’s Acosta, Rodriguez, Prieto-Valiente, Garcia-Segura, Wandosell and Alonso. This is an disease involve neuronal adenosine A2A receptors. Nat. Commun. 7:11915. open-access article distributed under the terms of the Creative Commons Attribution doi: 10.1038/ncomms11915 License (CC BY). The use, distribution or reproduction in other forums is permitted, Williard, D. E., Harmon, S. D., Kaduce, T. L., Preuss, M., Moore, S. A., provided the original author(s) and the copyright owner(s) are credited and that the Robbins, M. E. C., et al. (2001). Docosahexaenoic acid synthesis from n-3 original publication in this journal is cited, in accordance with accepted academic poluyunsaturated fatty acids in differentiated rat brain astrocytes. J. Lipid Res. practice. No use, distribution or reproduction is permitted which does not comply 42, 1368–1376. with these terms. Frontiers in Aging Neuroscience | www.frontiersin.org 15 December 2019 | Volume 11 | Article 346