NeuroMolecular Medicine (2021) 23:140–160 https://doi.org/10.1007/s12017-020-08639-7 ORIGINAL PAPER A Walnut Diet in Combination with Enriched Environment Improves Cognitive Function and Affects Lipid Metabolites in Brain and Liver of Aged NMRI Mice Carsten Esselun1 · Benjamin Dilberger1 · Carmina V. Silaidos1 · Elisabeth Koch2 · Nils Helge Schebb2 · Gunter P. Eckert1 Received: 31 August 2020 / Accepted: 25 November 2020 / Published online: 26 December 2020 © The Author(s) 2020 Abstract This in vivo study aimed to test if a diet enriched with 6% walnuts alone or in combination with physical activity supports healthy ageing by changing the oxylipin profile in brain and liver, improving motor function, cognition, and cerebral mito- chondrial function. Female NMRI mice were fed a 6% walnut diet starting at an age of 12 months for 24 weeks. One group was additionally maintained in an enriched environment, one group without intervention served as control. After three months, one additional control group of young mice (3 weeks old) was introduced. Motor and cognitive functions were measured using Open Field, Y-Maze, Rotarod and Passive Avoidance tests. Lipid metabolite profiles were determined using RP-LC-ESI(-)-MS/MS in brain and liver tissues of mice. Cerebral mitochondrial function was characterized by the determi- nation of ATP levels, mitochondrial membrane potential and mitochondrial respiration. Expression of genes involved with mito- and neurogenesis, inflammation, and synaptic plasticity were determined using qRT-PCR. A 6% walnut-enriched diet alone improved spatial memory in a Y-Maze alternation test (p < 0.05) in mice. Additional physical enrichment enhanced the significance, although the overall benefit was virtually identical. Instead, physical enrichment improved motor performance in a Rotarod experiment (p* < 0.05) which was unaffected by walnuts alone. Bioactive oxylipins like hydroxy-polyunsat- urated fatty acids (OH-PUFA) derived from linoleic acid (LA) were significantly increased in brain (p** < 0.01) and liver (p*** < 0.0001) compared to control mice, while OH-PUFA of α-linolenic acid (ALA) could only be detected in the brains of mice fed with walnuts. In the brain, walnuts combined with physical activity reduced arachidonic acid (ARA)-based oxylipin levels (p < 0.05). Effects of walnut lipids were not linked to mitochondrial function, as ATP production, mitochondrial membrane potential and mitochondrial respiration were unaffected. Furthermore, common markers for synaptic plasticity and neuronal growth, key genes in the regulation of cytoprotective response to oxidative stress and neuronal growth were unaffected. Taken together, walnuts change the oxylipin profile in liver and brain, which could have beneficial effects for healthy ageing, an effect that can be further enhanced with an active lifestyle. Further studies may focus on specific nutrient lipids that potentially provide preventive effects in the brain. Keywords Mitochondrial function · Brain · Liver · Oxylipins · Cognition · Behaviour · NMRI · Ageing · Neurodegeneration Supplementary Information The online version contains Abbreviations supplementary material available at https ://doi.org/10.1007/s1201 Aβ Amyloid-beta peptide 7-020-08639- 7. AD A lzheimer’s Disease ADP Adenosine diphosphate * Gunter P. Eckert eckert@uni-giessen.de ALA α -Linolenic acid, C18:3n3 ARA Arachidonic acid, C20:4n6 1 Laboratory for Nutrition in Prevention and Therapy, ATP A denosine triphosphate Institute of Nutritional Sciences, Justus-Liebig-University, COX C yclooxygenase Biomedical Research Center Seltersberg (BFS), Schubertstr. 81, 35392 Giessen, Germany CYP C ytochrome 2 P450 monooxygenase Chair of Food Chemistry, Faculty of Mathematics and Natural Sciences, University of Wuppertal, Gaussstr. 20, DBC Dissociated brain cells 42119 Wuppertal, Germany V1ol :.3(1234567890) NeuroMolecular Medicine (2021) 23:140–160 141 DGLA Dihomo-γ-linolenic acid, R123 Rhodamine 123 C20:3n6 ROS R eactive oxygen species DHA D ocosahexaenoic acid, C22:6n3 RP-LC-ESI(-)-MS/MS Reverse-phase liquid chromatog- DiHEDE Dihydroxy eicosadienoic acid raphy negative electrospray ioni- DiHETE Dihydroxy eicosatetraenoic acid zation tandem mass spectrometry DiHETrE Dihydroxy eicosatrienoic acid sEH Soluble epoxide hydrolase DiHDPE Dihydroxy docosapentaenoic SNP Sodium nitroprusside acid Tx Thromboxanes DiHODE Dihydroxy octadecadienoic acid Wal Walnut-only group DiHOME Dihydroxy octadecenoic acid WalEE Walnut + enriched environment DiOH-PUFA Dihydroxy polyunsaturated fatty group acid youCon Young control group EPA E icosapentaenoic acid, C20:5n3 EpDPE Epoxy docosapentaenoic acid EpEDE E poxy eicosadienoic acid Introduction EpETE E poxy eicosatetraenoic acid EpETrE Epoxy eicosatrienoic acid In physiological ageing as well as in neurodegenerative EpODE Epoxy octadecadienoic acid diseases like Alzheimer’s disease (AD) or Parkinson’s dis- EpOME E poxy octadecenoic acid ease, mitochondrial dysfunction (Swerdlow et al. 2014) Ep-PUFA Epoxy polyunsaturated fatty acid and chronic inflammation (Minciullo et al. 2016) appear ETC Electron transfer chain to play key roles in cognitive decline and decreased motor GC/MS G as chromatography/mass function. Hallmarks of mitochondrial dysfunction include spectrometry changes of the oxidative phosphorylation system (OXPHOS) HDHA Hydroxy docosahexaenoic acid leading to reduced complex activity, depolarization of the HEPE Hydroxy eicosapentaenoic acid mitochondrial matrix and inner membrane space and there- HETE Hydroxy eicosatetraenoic acid fore to a reduced ATP production (Grimm et al. 2016). HETrE Hydroxy eicosatrienoic acid Changes to complex-I and -III of the OXPHOS system also HODE Hydroxy octadecadienoic acid lead to increased reactive oxygen species (ROS) produc- HOTrE Hydroxy octadecatrienoic acid tion, potentially accelerating processes leading to neuronal LA Linoleic acid, C18:2n6 loss. Low-grade chronic, systemic inflammation during age- LLOQ Lower limit of quantification ing, often referred to as “inflammaging”, is based on the LN2 L iquid nitrogen body’s decreasing ability to ameliorate inflammatory events, LOD L imit of detection leading to an increased production of pro-inflammatory LOX Lipoxygenase cytokines like IL1β or TNF1α or prostaglandins such as MiR05 Mitochondrial respiration PGE2 or thromboxanes (Minciullo et al. 2016; Neves and medium Sousa-Victor 2019). Arachidonic acid (ARA) is predomi- MMP Mitochondrial membrane nantly metabolized by cyclooxygenases (COX), lipoxyge- potential nases (LOX), or cytochrome P450 monooxygenases (CYP) n3-PUFA Omega-3 polyunsaturated fatty to a diverse pattern of eicosanoids with a wide range of acid biological roles (see Fig. 1). Several products, particularly n6-PUFA Omega-6 polyunsaturated fatty prostaglandins as well as leukotrienes act pro-inflammatory acid (Nayeem 2018). The same enzymes also oxidize n3-fatty NMRI Naval Medical Research Institute acids like α-linolenic acid (ALA), eicosapentaenoic acid NO Nitric oxide (EPA) or docosahexaenoic acid (DHA). However, most OH-PUFA Hydroxy-polyunsaturated fatty of these metabolites are widely considered to exhibit anti- acid inflammatory effects (Nayeem 2018). Since both, n6- and oldCon Old control group n3-fatty acids (Kutzner et al. 2017) compete as substrates for OXPHOS Oxidative phosphorylation COX, LOX and CYP, dietary supplementation with specific PG Prostaglandins fatty acids is suspected to modify the metabolite profile in PI Protease inhibitor the body (Nayeem 2018; Ostermann et al. 2017a). PUFA P olyunsaturated fatty acid ALA is a precursor of long-chain n3-polyunsaturated qRT-PCR Q uantitative real-time polymer- fatty acids (n3-PUFA) like EPA and DHA and cannot be ase chain reaction produced by the body itself. Supplementation of ALA has 1 3 142 NeuroMolecular Medicine (2021) 23:140–160 shown to increase brain DHA levels (Eckert et al. 2010) This study aimed to investigate whether a 6% walnut- and anti-inflammatory oxylipins (Desai et al. 2016). Fur- enriched diet reflecting an uptake of 28 g/day for humans thermore, ALA was also found to reduce Aβ induced neuro- (Willis et al. 2009), could lead to a shift in the oxylipin inflammation and cognitive dysfunction in C57BL/6N mice profile and also affect cognition of mice by improving mito- (Ali et al. 2020). ALA also attenuated its toxicity in SH- chondrial function and neuronal growth. The study design SY5Y cells (Litwiniuk et al. 2020). It was further found to is based on our previous work investigating the effect of increase brain plasticity in C57BL/6N mice following ALA rice bran extracts in aged Naval Medical Research Institute injections (Blondeau et al. 2009) and reduce PD symptoms (NMRI) mice (Hagl et al. 2016). Old and young control in C. elegans model (Shashikumar et al. 2015). In general, mice were fed with a control diet mainly based on oleic acid. we and others reported that other n3-PUFA like DHA and Nuts and oleic acid have already been shown to positively EPA exhibit protective effects against neurodegenerative enhance cognition in elderly subjects (Martínez-Lapiscina diseases like PD (Bousquet et al. 2008), but also support et al. 2013) and benefit other parameters like inflammation cognitive development in infants (Dyall 2015; Weiser et al. or blood pressure (Sales-Campos et al. 2013). 2016) and be beneficial maintaining it during ageing (Cutuli In rats, this diet has already shown to improve cognition 2017). Rey et al. also reported that feeding of mice with i.e. by enhancing memory, but also improved animals’ motor a n3-PUFA rich diet still retained their anti-inflammatory function (Willis et al. 2009; Haider et al. 2011). In another oxylipin profile following LPS-induced inflammation (Rey study, Pandaresh et al. reported that a diet enriched with wal- et al. 2019) compared to n3-PUFA deficient mice. nuts affected ROS production and markers of oxidative stress Walnuts (juglans) consist of about 47% PUFA (Ros 2009) positively after 5–15 months of feeding in a murine model of and have one of the highest concentration of ALA of all AD (Pandareesh et al. 2018). Oxidative and nitrosative stress edible nuts (Carey et al. 2013). Favourably, they also have are a hallmark not only of AD, but also of the physiologi- a high ratio of n3-PUFA/n6-PUFA of 1:3 to 1:4 (Vinson cal ageing process (Dilberger et al. 2019; Jiménez-Jiménez and Cai 2012; Poulose et al. 2014). n3-PUFA are known to et al. 2016). A study by Liu et al. found that oral gavage reduce the production of peptide amyloid-beta (Aβ) (Emen- of 600 mg/(kg*d) of walnut kernels would protect mice in dato et al. 2016; Thomas et al. 2015), commonly associ- an ageing model using d-galactose-induced liver and brain ated with AD, and additionally improve health of brain damages (Liu et al. 2019), proposedly by enhanced ATP cells (Carey et al. 2013). For example, in accordance with production and normalized acetylcholinesterase activity. Cole et al. we reported that DHA improved the fluidity of Currently, data regarding synergistic effects of exercise and neuronal membranes (Cole et al. 2009; Eckert et al. 2011). PUFAs, in form of walnuts, regarding oxylipin metabolism, Furthermore, n3-PUFA and their metabolites modulate gene cognitive function and mitochondrial performance are very expression in the brain including genes involved in mito- limited. In this regard, this study might provide valuable chondrial biogenesis, ATP production and oxidative metabo- insights and a foundation for further research in this area on lism (Eckert et al. 2013). the effect of walnuts, exercise and ageing. Intake of walnuts was associated with improved working memory in elderly people (Valls-Pedret et al. 2012,2015). Besides PUFAs, walnuts are rich in polyphenols and vitamin Methods E (McKay et al. 2010; Bourre 2006a, b ), which are also important for a healthy brain, as their antioxidative proper- Chemicals ties are an intrinsic counteraction to n3-PUFAs susceptibility to lipid peroxidation (Carey et al. 2013; Cole et al. 2009; All chemicals used for this research were of highest purity Rosales-Martínez et al. 2017) or their effect on mitochondria available and purchased from either Sigma Aldrich, Merck affected survival in mice and nematodes (Dilberger et al. or VWR. Oxylipin standards and internal standards were 2019). acquired and prepared as previously described (Kutzner Several studies have shown that a physically active life- et al. 2019; Rund et al. 2018; Ostermann et al. 2020; Koch style encouraged by an enriched environment attenuated et al. 2020). Aqueous solutions were prepared with type-1 cognitive decline in ageing and AD models (Maesako et al. ultrapure water. 2012; Costa et al. 2007; Baraldi et al. 2013; Verret et al. 2013) and improves mitochondrial function in the brain Animals (Schaffer et al. 2012; Steiner et al. 1985; Lores-Arnaiz et al. 2010; Asseburg et al. 2016). Moreover, acute and chronic Female NMRI (Naval Medical Research Institute) mice were exercise mobilized oxylipins that are most related to inflam- acquired from Charles River (Sulzbach, Germany) and kept matory processes, tissue repair or oxidative stress (Signini in the animal facility of the pharmacological institute of the et al. 2020). Goethe University Frankfurt am Main until they reached the 1 3 NeuroMolecular Medicine (2021) 23:140–160 143 Table 1 General components of C1000 diet. Modification of diets 15, 3-week-old mice was added 3 months after the start of were based on fat composition (according to the manufacturer) the feeding period to ensure that feeding periods ended at General Components Control diet (sunflower 6% walnut- the same time point. All mice had ad libitum access to diets oil) enriched and water and the young control group received the same [%] diet diet as the aged control group. Behavioural testing was per- [%] formed again at the end of the feeding period before mice Protein 17 18 were sacrificed via decapitation. Cerebellum, brain stem and Fat 5.7 5.8 olfactory bulb were removed from the brain before it was Carbohydrates 47 45 dissected on ice for further experimentations. Liver tissue Ash 5.5 5.9 was snap frozen after dissection. Fibre 3.1 3.1 Sugar 11 11 Determination of Fat Composition Vitamin E [µg] 180 160 and Quantification Fatty acid composition and quantification was carried out Table 2 Detailed fat composition of each diet. Fatty acid content was following a protocol of Weibull-Stoldt’s according to AOAC determined as methyl esters by gas chromatography-mass-spectrome- 963.15 (Official Methods of Analysis of AOAC Interna- try after lipid extraction and derivatization (see material & methods) tional. 1995). In brief, diet pellets were finely grounded in Fatty acid Control diet (sunflower 6% walnut- liquid nitrogen and hydrolysed in HCl for 30 min. Fatty acid oil) enriched diet extraction was performed using a Soxhlet Extraction System [%] [%] (Gerhardt Analytical Systems, Königswinter, Germany) in Palmitic acid 3.37 7.71 light petroleum. Stearic acid 1.12 1.85 Prior to analysis via GC/MS, fatty acids were methyl- Oleic acid 89.0 44.2 ated with B F3 (20% in MeOH). Before injection, GC/MS Vaccenic acid 1.62 1.42 samples were further diluted 1:10 with isooctane. Only Linoleic acid 4.90 39.5 peaks > 0.05% of the highest peak found were considered Linolenic acid – 5.35 for analysis. NIST Mass Spectral Library 2.0 g was used for Total fats 5.70 5.80 identification of compounds. Open Field starting age of 12 months. Female mice were chosen because of later re-housing, which could have resulted in increased For open field experiments, mice were placed in the mid- injury of male mice as they tend to show increased rivalry dle of a 45 × 45 cm big arena. Mice were then allowed to in a new environment. Initial Y-maze and Rotarod tests were freely roam around in the area for 5 min. Mice were filmed conducted in order to divide all mice into 3 homogenous via a camera placed above the arena pointing downwards groups à 15 mice. After that, mice were moved into new and recorded with TSE VideoMot 3D Classic V8.02. Later, cages in order for all groups to be of equal skill at the start videos were blinded by a second party, reformatted sev- of the trial. One group (oldCon) received a control diet (con- eral times using DaVinci Resolve 15 (Blackmagic Design) trol diet C1000, Altromin, Lage, Germany) modified with and finally analysed using MouseMove Version 1, an open sun flower oil (Lamotte OILS, Bremen, Germany) as basis source program used to analyse the movement of mice (Sam- for fats resulting in a total amount of 5.8% fat, for 6 months son et al. 2015). The composition of the diet including the fatty acid pat- Data for travel distance, speed, left and right turns stop tern are shown in Tables 1 and 2. Two groups (Wal and fraction and number of faecal droppings were analysed WalEE) received an identical pelleted diet enriched with 6% using GraphPad Prism version 8.2. for Windows (Graph- walnuts (Walnut, 60 g/kg diet) (provided by the California Pad Software). Walnut Commission, California, USA) for 6 months. Total fat amount of walnut diet was 5.8% and consisted of mainly Y‑Maze oleic acid, n6-PUFAs LA and the n3-PUFA ALA. A detailed list of all fatty acids contained in the diets can be found Cages were blinded by a second party, before the start in Table 2. Additionally, one walnut group (WalEE), were of the experiment. Mice were placed in one arm of a housed in bigger rat cages with an enriched environment in Y-shaped maze. On the intersection of all three arms, vis- form of running wheels, houses and other objects encour- ual indicators in form of shapes were placed in order for aging physical activity. A young control group (youCon) of the mouse to orient itself. A mouse was placed into one 1 3 144 NeuroMolecular Medicine (2021) 23:140–160 randomly chosen arm of the maze and left to explore the would remember the stimulus, it would generally stay in maze for 5 min without further disturbance of the experi- the bright chamber for a longer period of time. menter. Entry into a new arm, labelled A, B or C, was doc- umented. Later the number of entries was determined as Preparation of Dissociated Brain Cells From Freshly well as the order in which a mouse traversed the maze. A Isolated Brains full alternation was defined as a mouse visiting all 3 (e.g. A → B → C) arms, before returning to an already visited Dissociated brain cells (DBC) were prepared from one brain one. If the mouse entered arm A next (A → B → C → A), hemisphere according to a previously published protocol a second alternation would be counted (A → B → C & (Franke et al. 2007). For MMP measurements, DBCs were B → C → A) Alternation rate is given as ((number of alter- diluted in un-supplemented DMEM (Gibco, Thermo Sci- nations/total number of possible alternations) * 100). entific) and seeded into 24-well plates (250 µL per well). For ATP measurements, the diluted DBCs were seeded in 96-well plates (50 µL per well). Additionally to basal MMP Rotarod and ATP levels, a subset of DBCs was also treated with sodium nitroprusside (SNP) (2 mM for MMP and 0.1 mM The Rotarod (Accelerating Rota Rod, Panlab/Harvard for ATP) to simulate increased nitrosative stress found in Apparatus) was set to a base speed of 13 rpm. The speed ageing. DBCs were than incubated at 37 °C and 5% C O2 for increased over the course of one min to a maximum speed 3 h before measurement. of 40 rpm. At base speed, mice were placed on the rod and the timer was started. The timer was stopped if a mouse Mitochondrial Membrane Potential (MMP) fell from the rotating rod onto a pressure plate. If a mouse did not fall from the rod after 2 min, the test was also MMP was measured using fluorescence dye rhodamine 123 stopped. (R123). Dissociated brain cells of freshly dissected brain Every mouse was trained 3 times for 2 consecutive were incubated at 37 °C and 5% CO2 for 15 min with 0.4 µM days. Mice had a 15 min break between each training run. R123. Cells were than centrifuged at 750 g for 5 min and The final test was performed on the third day. The latency washed with HBSS buffer (supplemented with Mg2+, C a2+ to fall was calculated as the mean time to fall from 3 runs and HEPES; pH 7.4; 37 °C). Cells were resuspended in fresh on the final day of testing. Experimenter was blinded to HBSS buffer before R123 fluorescence was determined. The which mice he was testing. excitation wavelength was set to 490 nm and the emission wavelength to 535 nm on a ClarioStar plate reader (BMG Labtech, Ortenberg, Germany). The fluorescence was meas- Passive Avoidance ured 4 times and normalized to protein contents assessed via BCA method. In order to investigate the mice’s ability to remember a negative event, a passive avoidance experiment was per- ATP Levels formed. The passive avoidance chamber (Passive Avoid- ance Step-trough New Model, Ugo basile) consists of two ATP concentrations were determined using an ATPlite chambers. One chamber is illuminated by a bright light Luminescence Assay System (Perkin Elmer, Rodgau- of 1,350 lm, the other chamber is dark. Both chambers Jügesheim, Germany), which is based on the light emission are separated by a wall with a door. On the first day of of luciferin upon reaction with ATP. The 96-well plate was the experiment, a mouse was placed in the bright cham- removed from the incubator and allowed to cool to room ber. Shortly after, the door to the dark chamber is opened temperature for 10 min. Following incubation with lysis and the time of how long it took the mouse to completely buffer for 10 min in the dark the monitoring reagent was enter the dark chamber was recorded. Once it was inside, added to the wells. The plate was again incubated in a dark the door closed and a small electrical stimulus of 0.5 mA environment at room temperature for 5 min. The emitted was applied to the mouse. The mouse was then removed light was assessed with a ClarioStar plate reader (BMG from the apparatus and placed back into its cage. On the Labtech, Ortenberg, Germany). The results were adjusted second day, the mouse was again placed into the chamber. to protein content. This time, the door to the dark chamber was opened from the start. Upon entering the dark chamber there would be no electrical stimulus. The time it took for the mouse to enter the dark chamber was recorded. As a healthy mouse 1 3 NeuroMolecular Medicine (2021) 23:140–160 145 Isolation of Brain Mitochondria and High‑Resolution 0.5 mM oxaloacetate, 50 µM EDTA, 0.31 mM acetyl coen- Respirometry zyme A, 5 mM triethanolamine hydrochloride and 0.1 M Tris–HCl) was mixed and heated to 30 °C for 5 min. A Frontal half of one brain hemisphere was homogenized in volume of 10 µL of the mitochondria sample was added to 2 mL MiR05 (mitochondrial respiration medium developed the reaction medium and the citrate synthase activity was by Oroboros (Gnaiger 2020)) containing EGTA (0.5 mM), determined photospectrometrically at 412 nm. Measure- magnesium dichloride (3 mM), lactobionic acid (60 mM), ments were performed in triplicate. taurine (20 mM), potassium dihydrogen phosphate (10 mM), HEPES (20 mM), sucrose (110 mM) and essentially fatty Protein Content acid free bovine serum albumin (1 g/L). Sample was than centrifuged at 1.400 g for 7 min at 4 °C to remove cell A small sample of mitochondria from the respirometry debris. Supernatant was centrifuged at the same conditions measurements or DBCs were frozen in L N2 and stored for 3 min, than at 10,000 g for 5 min and 4 °C to yield a pel- at -80 °C until assessment of protein contents. Samples let of crude mitochondria. For further isolation, a protease were thawed and protein contents were determined using inhibitor (PI, complete tablets, Roche, Mannheim, Germany) Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, was added to the medium. The mitochondrial pellet was Waltham, MA, USA) according to manufacturer’s instruc- resolved in MiR05 + PI and centrifuged again at 1,400 g for tions. Absorbance was measured using a ClarioStar plate 3 min at 4 °C. Supernatant was again centrifuged at 10,000 g reader (BMG Labtech, Ortenberg, Germany). for 5 min at 4 °C to yield the pure mitochondria which were than resolved in 900 µL MiR05 + PI of which 80 µL were Quantitative Real‑Time PCR (qRT‑PCR) injected into the Oxygraph for high-resolution respirometry. Investigation of the respiratory system was performed Freshly dissected brain tissue (~ 20 mg) from the rear part of following a protocol created by Gnaiger et al. (Gnaiger a brain hemisphere was stabilized with RNAlater (Qiagen, 2020). After injection of mitochondria into the Oxygraph Hilden, Germany), frozen in LN2 and stored at -80 °C until 2 k chamber (Oroboros, Innsbruck, Austria), complex-I experimentation. Total RNA was isolated from brain tis- (CI) activity was determined by addition of 5 mM pyruvate sue using the RNeasy Mini Kit (Qiagen, Hilden, Germany) and 2 mM malate (CIL). Addition of 2 mM ADP reflected according to the manufacturer’s instructions. physiological CIP activity. Subsequent addition of 10 mM Assessment of purity and quantification was achieved succinate activated complex-II, resulting in full physi- using a NanoDrop 2000 × spectrophotometer (Thermo Fisher ological respiration CI + IIP. Mitochondrial integrity was Scientific, Waltham, MA, USA). RNA was treated with examined via addition of 10 µM cytochrome C. Uncoupled TurboDNA free Kit (Qiagen, Hilden, Germany) to further CI + IIE was determined via stepwise addition of carbonyl improve quality of isolated RNA. cDNA was synthesized cyanide p-(trifluoromethoxy) phenylhydrazone (FCCP, up using an iScript cDNA synthesis kit (BioRad, Munich, Ger- to 1–1.5 µM). Uncoupled CIIE was measured following the many) according to manufacturer’s instructions. To perform addition of complex-I inhibitor rotenone (0.5 µM). Leak qRT-PCR, SYBR Green technology was used on a CFX96 respiration (CIIL) was achieved by addition of oligomy- Touch real-time PCR detection system (BioRad, Munich, cin (2 µg/mL) and residual oxygen consumption, caused Germany). qRT-PCR was performed in a total volume of by enzymes which are not involved in the electron transfer 10 µL assessing each sample in triplicate. Primer sequences, chain, was determined via by addition of 2.5 µM antimy- concentrations, product sizes and PCR conditions can be cin A. Finally, C IVE activity was determined by addition of found in Table 3. Gene expression was analysed using the 0.5 mM tetramethylphenylenediamine (TMPD), an artificial 2(-∆∆Cq) method with a BioRad CFX manager (BioRad, substrate of complex IV and 2 mM ascorbate to regener- Munich, Germany) and normalized to phosphoglycerate ate TMPD. All data points were corrected for residual oxy- kinase 1 (PGK1) and beta-2-microglobulin (B2M) expres- gen consumption and CIVE was additionally corrected for sion levels. All used primers for qPCR were acquired from autooxidation rate, determined via addition of excess NaN3 Biomol (Hamburg, Germany). after the final step of the experiment. Determination of Free Oxylipin Profile in Brain Citrate Synthase Activity and Liver A small sample of mitochondria from the respirometry Free oxylipins were analysed as previously described measurements was frozen in L N2 and stored at − 80 °C until (Kutzner et al. 2019; Rund et al. 2018,2019). In brief, 10 µL assessment of the citrate synthases activity. Samples were of internal standards, butylated hydroxy toluene and enzyme allowed to thaw while a reaction medium (0.1 mM DTNB, inhibitors were added to ~ 20–50 mg tissue. The samples 1 3 146 NeuroMolecular Medicine (2021) 23:140–160 were homogenized in 300 µL methanol using a ball mill. Mice were excluded from the experiments if they did Samples were centrifuged, the supernatants were collected not appear to be of sufficient health for behavioural testing. and loaded onto the pre-conditioned solid phase extraction Outliers were removed according to ROUT’s outlier test (SPE) cartridges. Oxylipins were extracted by SPE using a applying Q = 1%. For oxylipin analysis, statistical evalu- C8/anion exchange cartridge material (Bond Elut Certify II, ation was done if > 50% of samples met the lower limit Agilent, Waldbronn, Germany). The eluate was evaporated of quantification (> LLOQ) and if the mean of all sam- and reconstituted in methanol containing second internal ples was still > LLOQ. All samples which did not reach standards. Samples were analysed by means of LC–MS/MS the LLOQ were set to their specific limit of detection (QTRAP, Sciex, Darmstadt, Germany) with negative elec- (LOD) for calculation. If the concentration of an oxylipin trospray ionization. Detection was carried out in scheduled was < LLOQ for > 50% of the group’s mice, the analysed selected reaction monitoring mode. group was excluded from analysis and “ < LLOQ” is dis- played instead. Data Handling and Statistics Unless stated otherwise, data are presented as mean ± SEM. Statistical analysis was performed using either a student’s t-test or an one-way ANOVA followed by Dunnett’s post-test performed with GraphPad Prism version 8.2 for Windows (GraphPad Software). Fig. 1 Simplified PUFA synthesis and origin of oxylipins analysed left, n3-PUFA on the right. The oxylipins act as potent lipid media- within this study formed by cyclooxygenase (COX), lipoxygenase tors showing diverse biological activity ranging from pro-inflam- (LOX) and cytochrome P450  monooxygenase (CYP) (Gabbs 2015; matory (e.g. PG, LT) to analgesic, vasodilatory, anti-inflammatory Ostermann et al. 2017b; Coras 2020). Bold arrows connecting PUFA action (Epoxy-PUFA). Oxylipins from n3-PUFA are commonly indicate a multiple-step elongation process between each of them. considered to be less potent inflammatory mediators compared to Thin arrows indicate oxylipins formed by the exemplary indicated n6-PUFA oxylipins or even anti-inflammatory. Displayed are only a enzymes from the PUFA in single or multiple-step reactions. It few selected oxylipins, not a complete set of all oxylipins known to should be noted that several of the oxylipins can also be formed by be produced. Data for underlined oxylipins can be found in Table 5 other pathways as well as autooxidation. n6-PUFA are shown on the or Fig. 5 1 3 NeuroMolecular Medicine (2021) 23:140–160 147 Results stressful events in a measurable way. Results for Open Field experiments and faecal droppings can be found in supple- Treatment mentary Fig. S2 and supplementary Table S1 and supple- mentary Table S3. After reaching adulthood the mice’s bodyweights stayed stable for the whole feeding period. Mice in each group Mitochondrial Parameters showed a similar range of bodyweights, although they dif- fered by around 15–20 g inside of each group. Weight Complex activities of the oxidative phosphorylation system changes for mean bodyweights and individual bodyweights of isolated mitochondria from freshly dissected mice brain can be found in supplementary Table S10. Mice from both showed no differences between the groups, indicating that Wal and WalEE groups were around 1.5–2 g heavier than the interventions did not have an effect on mitochondrial the oldCon mice during the whole feeding period. respiration (Fig. 3a). Only a trend for WalEE mice to have reduced complex activity could be found. As seen in Fig. 3f, Behavioural Tests using citrate synthases activity as a marker for mitochondrial mass, mitochondrial content also seemed to be unaffected Aged mice showed significant deficits in Y-Maze alternation by the interventions. (p*** = 0.0009) and Rotarod tests (p**** < 0.0001) com- Furthermore, MMP (Fig. 3b), which was determined as pared to young mice (Fig. 2a–d). A 6% walnut-enriched diet R123 fluorescence, was unaffected by the interventions, showed significantly higher numbers of entries in Y-Maze supporting the next to identical complex activity of the test (p* = 0.0237), but did not have an effect on motor func- OXPHOS. Sodium nitroprusside (SNP) was used to induce tion in Rotarod test (p = 0.9999). Number of alternations nitrosative stress to the DBCs as the generation of reactive (Wal: p* = 0.0157; WalEE: p*** = 0.0007) and alternation nitrogen species (RNS) is a common process in ageing. After rate (Wal: p* = 0.0207; WalEE: p* = 0.0421) were signifi- addition of 0.5 mM SNP (Fig. 3c) and incubation for 3 h, cantly increased in both Wal and WalEE. Although the inter- a decrease in MMP of approximately 15% was observed. vention group with additional physical enrichment showed Neither interventions, however, lead to an improved reaction the same results, they were more homogenous among all to SNP-insult. mice leading to a higher significance in the number of Although ATP is mainly produced in the OXPHOS, a alternations (p*** = 0.0007) but not for the alternation rate minor part of ATP is also generated, for example, during (p* = 0.0421). While motor functions in Wal mice were vir- glycolysis. Since respiration and MMP were unaffected, we tually identical to aged control mice, physical enrichment measured ATP levels. As displayed in Fig. 3d, ATP levels attenuated the age-dependant decrease in motor function were not increased in either intervention group. After insult significantly (p* = 0.0381). with SNP (Fig. 3e), ATP levels of all groups were reduced by Generally all groups in the Passive Avoidance test around 40% to 33% compared to untreated DBCs, whereas (Fig. 2e) showed a higher delay before entering the dark walnut interventions or additional physical enrichment had chamber on the second day suggesting that mice remem- only little to no effect on the induced damages (Table 3) bered the adverse event. However, only trends could be identified by statistical evaluation. Walnut intervention Gene Expression in the Brain alone did not appear to affect passive avoidance performance (p = 0.9716), but physically active mice tended to stay in To gain further mechanistic insight we investigated the gene the illuminated chamber longer than aged mice (p = 0.1507). expression of several markers for neuronal growth, synap- Young control mice only showed virtually identical behav- tic plasticity and antioxidant system (Table 4). CREB1, iour like aged mice (p = 0.9930). which is closely related to long time memory, was signifi- We also performed an open field experiment to exam- cantly decreased in aged mice compared to young mice ine the general behaviour and activity of mice. We assessed (p* = 0.0149), WalEE was virtually identical with oldCon mice’s travelled distance, stop fraction, average speed and (p = 0.8025). In general, gene expression in aged mice turn. We found only little numerical differences in all meas- showed trends to be decreased for all investigated genes, ured parameters. Aged control mice showed a small trend indicating possible reduced synaptic plasticity (Synaptophy- to underperform, while other groups were virtually iden- sin, GAP43), neuronal growth (NGF, BDNF) or response tical. Quantity of faecal droppings during the Open Field to oxidative stress (KEAP1, NRF2). Neither walnut-only experiment is often linked to the emotionality of the animals nor WalEE interventions were able to attenuate this age- (Seibenhener and Wooten 2015). Our results suggest that the dependant reduction. interventions did not affect mice’s sensitivity to emotional/ 1 3 148 NeuroMolecular Medicine (2021) 23:140–160 Fig. 2 Y-Maze alternation test. a Number of entries into the three chamber. On day one, a weak stimulus was applied upon enter- different arms of the Y-Maze. Duration of Y-Maze experiment was ing the dark chamber. On the second day, time until mice entered 5 min. b Number of full alternations. One full alternation is defined the dark chamber was recorded to investigate mice’s memory. Dis- as a mouse entering all arms of the maze before re-entering the played are mean ± SEM of n = 12–15 in a–d and n = 10–12 in e. already visited one. c Alternation rate of Y-Maze test, calculated as Full range of data is displayed using dots for each individual mice. the number of full alternations divided by the number of entries. d Statistical significance was tested via one-way ANOVA and post Rotarod test. Time in [s] of mice being able to stay on the rotating hoc Dunnett’s test comparing all groups with oldCon (*p < 0.5, rod. Speed of the rod gradually increased until 60 s into the experi- **p < 0.01, ***p < 0.001, ****p < 0.0001). oldCon = aged Con- ment; experiment was stopped if mice were able to stay on the rod trol; youCon = Young control; Wal = Walnut group; WalEE = Wal- for 120 s. e Passive Avoidance test. Mice were put in an illuminated nut + Enriched Environment group. Parameters for the statistical tests chamber and time was measured until they retreated into a dark can be found in the supplementary Table S1 1 3 NeuroMolecular Medicine (2021) 23:140–160 149 Oxylipin Profile and Gene Expression in the Liver nor mitochondrial function were affected by the intervention diet, we looked at whether the diet was affecting fatty acid As our data suggest that neither gene expression in the brain metabolism. For this reason, and to investigate whether the 1 3 1 50 NeuroMolecular Medicine (2021) 23:140–160 ◂Fig. 3 a Oxygen consumption of mitochondria isolated from the In general, across all groups, ARA-based prostanoid brain adjusted to protein. Activity of OXPHOS complexes were levels are around 80-fold higher in the brain tissue com- assessed via addition of several substrates, inhibitors or uncou- plers. Which substance was added in which stage of the experiment pared to the liver. As seen in Fig. 5, in the liver, the Wal is marked with “ + ”; n = 12. b Basal MMP levels were measured as group showed a general trend for reduced ARA-based pro- R123 fluorescence after incubation of DBC samples. c MMP levels inflammatory prostanoids to around the level of the young after 0.5  mM SNP-induced nitrosative stress. d Basal ATP concen- control group (if detectable). In the brain, the oldCon group tration measured in µM per mg protein in the DBC samples. Meas- ured signal stems from bioluminescence reaction of luciferin and showed a trend for higher prostanoid concentrations com- ATP. e ATP levels after SNP-induced nitrosative stress. DBCs have pared to the youCon mice. While Wal mice did not change been incubated for 3 h prior to measurement. f Citrate synthase activ- prostanoid levels, the WalEE group showed generally lower ity as an indicator of mitochondrial content. Amount of protein was concentrations akin to those of the young mice. Still, these measured via BCA method. Displayed are means ± SEM. Statistical significance was tested via one-way ANOVA with Dunnett’s post hoc changes were only a trend in the case of PGD2, but signifi- test using oldCon as reference for statistical comparisons. Each group cant for P GE2 (p* < 0.05), 6-ketoPGF1α (p* < 0.05), PGF2α in C-F consisted of 12–15 female NMRI mice. Parameters for the sta- (p*** < 0.001) and TxB2 (p*** < 0.001) compared to old- tistical tests can be found in the supplementary Table S2 Con. Other prostanoids were virtually identical between each group. The effect of physical enrichment also varied diets might have an effect on inflammatory processes, com- between the organs. In the liver, WalEE showed similar pros- monly increasing during the process of ageing, we explored tanoid levels compared to old control, with the exception the free, i.e. nonesterified oxylipin profiles in brain and liver being PGF2α, which was increased compared to old control tissue. (p** < 0.01). Prostanoid levels in the Wal group, however, As seen in Table 5 and Fig. 4 as well as in Fig. 5, the tended to be more in line with youCon. diet’s effect varied between the two organs. Table 5 shows Displayed in Fig. 4 is a heatmap showing increases or that in the liver hydroxy-PUFA (OH-PUFA) of walnut- decreases of several OH- and Ep-PUFAs derived from LA, related fatty acids were significantly increased compared to ALA, ARA, EPA and DHA relative to results of the oldCon the aged control group (p*** < 0.001 to p**** < 0.0001). group. Statistical evaluation of the analytes can be found The same was observed for epoxy-PUFA (Ep-PUFA), in supplementary Table S5 for brain and supplementary as both intervention groups showed concentrations of Table S6 for the liver. Results for the liver show that in a 50.5 ± 10 nmol/kg (Wal) and 54.6 ± 6.7 nmol/kg (WalEE), similar vein as for the ARA-based prostanoids, an enriched while the concentration in oldCon was < LLOQ for all environment in the WalEE group led to an all-encompassing measured EpODEs. Similar results were found for LA increase of oxylipins compared to oldCon. The Wal group (p* < 0.05 to p** < 0.01) and ALA (control < LLOQ)- on the other hand showed results more akin to the you- based oxylipins in the brain. Interestingly, in the brain Con group, whose ARA-derived oxylipins were reduced both intervention’s EPA and DHA OH-PUFA and Ep- compared to oldCon. Looking at data from brain tissue PUFA were virtually identical compared to both control (see Table 5), we found that both intervention groups had groups. Concerning the liver’s Ep-PUFA it is to note that reduced ARA-based oxylipins, while at the same time levels for ALA, mainly 15(16)-EpODE could be detected in of oxylipins derived from LA were increased. Due to high the intervention groups. The same could be said for their biological variability of the outbred NMRI mice, this cannot hydrolysis products, the DiHODEs, which can be formed be ambiguously said for the liver. by soluble epoxide hydrolase (sEH). Ep-PUFAs derived All changes considered, when looking at the n6/ from EPA could only be found in intervention groups, but n3-PUFA-oxylipin ratio in Table 6, it is interesting to note not in the control groups. The same is true for Ep-PUFA that in the brain only trends for a changed ratio could be of DHA with the exception of 10(11)-EpDPE, 13(14)- found, while the ratio in the liver samples was significantly EpDPE and 19(20)-EpDPE, which could also be found different (p**** < 0.0001). Part of the reason for this is in the control group. In the brain, Ep-PUFAs of LA and the high ARA-derived oxylipin concentration in the brain, EPA were < LLOQ in all groups and Ep-PUFAs of ALA, which overshadowed smaller changes in ALA, EPA or DHA specifically 15(16)-EpODE, could only be detected in the oxylipins. intervention groups. Looking at the expression of genes related to antioxida- As seen in Table 5, in the brain, OH-PUFA of ARA tive capacity in the liver (Table 7), we see similar results as were significantly reduced in the groups that were fed the in the brain (Table 4). It is to note that Wal mice showed a walnut-only diet (p** < 0.01). This change appears to origi- significantly higher expression of KEAP1 gene, which is nate mainly from the substantial reduction of 12-HETE. part of the KEAP1-NRF2 pathway of handling oxidative This effect was further enhanced in physically active mice stress. NRF2, however, did not show an increase, but tended (p*** < 0.001). The same trends can be found for brain Ep- to be decreased in both interventions groups. However, the PUFAs, but neither for OH-PUFAs or EP-PUFAs in the liver. expression of Interleukin 1β as a marker for inflammation 1 3 NeuroMolecular Medicine (2021) 23:140–160 151 Table 3 Primer sequences, Primer Sequence Size conc Annealing Temp manufacturer’s, product sizes [bp] [µM] (time) ( no. of cycles) (bp), concentrations (conc.) and programme used for qRT-PCR PGK1 5′-GCAG AT TGT TTG GAA TGG TC-3’ 185 0.4 58 °C (45 s), (45x) measurement 5′-TGC TCAC AT GGC TGA CTTT A-3’ B2M 5′-GGC CTG TAT GCT ATCC AGA A-3’ 198 0.4 58 °C (45 s), (45x) 5′-GAA AGA CCA GTC CTT GCG A-3’ BDNF 5′-GAT GCC AGT TGC TTT GTC TT-3’ 137 0.1 58 °C (45 s), (42x) 5′-ATG TGA GAA GTT CGG CTTT G-3’ NGF 5′-ATCA AG GGC AAGG AG GTG ACAG- ‘3 143 0.15 62 °C (20 s), (40x) 5′-GAGT TC CAG TGTT TG GAGT CG ATG -3 ‘ NRF2 5′-GAT CTCC TC GCT GGA AAA AG-3’ 187 0.4 56 °C (30 s), (50x) 5′-GTCA CTG GGC TC TGCT ATG A-3’ CREB1 5′-TAG CTG TGAC TT GGC ATT CA-3’ 184 0.5 58 °C (45 s), (45x) 5′-TTGT TCT GTT TGG GA CCTG T-3’ Synaptophysin 5′-TTT GTG GTTG TTG AG TTC CT-3’ 204 0.1 58 °C (45 s), (42x) 5′-GCAT TT CCT CCC CAA AGTA T-3’ Gap43 5′-AGG GAGA TG GCT CTGC TA CT-3’ 190 0.15 58 °C (45 s), (42x) 5′-GAG GAC GGG GAGT TA TCA GT-3’ Keap1 5′-ATG GCC ACAC TT TTCT GG AC-3’ 131 0.2 60 °C (45 s), (45x) 5′-TCC TGT TGTC AG TGCT CA GG-3’ IL1β 5′-CCC AAC TGG TAC ATCA GC AC-3’ 180 0.3 58 °C (30 s), (45x) 5′-TCT GCT CAT TCA CGAA AAG G-3’ showed a greater variability in aged control mice resulting Discussion in a trend for a higher expression in the liver, which could potentially be reduced by the interventions. In the brain, the Diet and Body Weight overall expression appears to be similar across all groups. A 6% walnut supplementation, which corresponds to 28 g/d for humans (Willis et al. 2009), has previously enhanced cognitive and motor function in humans and rodents (Wil- lis et al. 2009; Haider et al. 2011). Mice adhered to the diets during the whole feeding period and gained only a little weight at the beginning. Starting at around week 3, Relative normalized mRNA expression of relevant genes in mice kept their weight. As the NMRI mice were already Table 4 brain tissue. Data are adjusted to aged control group = 100%; n = 10; 12 months old at the start of the feeding trial, weight dif- Displayed are means ± SEM; one-way ANOVA with Dunnett´s multi- ferences and differences in their abilities were expected and ple comparison post hoc test using oldCon as reference for statistical adjusted for. Muthaiyah et al. who fed a 6% walnut-enriched comparison; normalized to the mRNA expression levels of beta-2-mi- croglobulin (B2M) and phosphoglycerate kinase 1 (PGK1) diet to tg-AD mice also showed no to little differences in the body weight during the first 4–9 months of their diet Gene OldCon YouCon Wal WalEE (Muthaiyah et al. 2014). We therefore propose that the diet [%] [%] [%] [%] was generally well received and adhered to. For cognitive Synaptophysin 100 ± 7.6 124 ± 8.0 99.3 ± 8.6 110 ± 6.4 and motor testing, as well as metabolic changes or mitochon- NGF 100 ± 9.4 145 ± 17 97.9 ± 18 115 ± 13 drial function, estrogenic state of female mice might affect BDNF 100 ± 27 228 ± 53 98.3 ± 28 115 ± 28 the variability of data. As behavioural tests and sacrifice was GAP43 100 ± 28 111 ± 18 86.6 ± 12 81.7 ± 13 conducted over a longer period of time, every state of the KEAP1 100 ± 11 144 ± 21 112 ± 18 124 ± 14 estrogenic cycle should be represented in the dataset. Fur- NRF2 100 ± 13 152 ± 25 89.5 ± 14 120 ± 23 thermore, recent studies analysing data from multiple studies CREB1 100 ± 8.7 144 ± 9.1 78.5 ± 15 111 ± 6.7 revealed that differences in variability between male and * p = 0.0149 female mice are negligible (Prendergast et al. 2014; Smarr IL1β 100 ± 17 116 ± 20 101 ± 20 158 ± 36 et al. 2019). Significance is displayed as: *p < 0.05 (oldCon vs youCon). Param- eters for the statistical tests can be found in the supplementary Table S8 1 3 1 52 NeuroMolecular Medicine (2021) 23:140–160 Table 5 Comparison of hydroxy- and epoxy-PUFA concentrations based on LA, ALA, ARA, EPA and DHA; N = 15; Displayed are in liver and brain and between all different intervention groups. A: means ± SEM of each groups. Data were statistically compared via an Sum of all brain hydroxy- and epoxy-PUFAs based on LA, ALA, one-way ANOVA with Dunnett’s post hoc test and oldCon group as ARA, EPA and DHA. B: Sum of all liver hydroxy- and epoxy-PUFAs reference (ref.) A Hydroxy-PUFA [nmol/kg] Epoxy-PUFA [nmol/kg] Brain derived from derived from LA ALA ARA EPA DHA LA ALA ARA EPA DHA oldCon 50.3 ± 6.7 < LLOQ 1480 ± 80 4.95 ± 1.0 145 ± 16 < LLOQ < LLOQ 40.6 ± 5.5 < LLOQ 12.3 ± 1.7 ref ref ref YouCon 47.4 ± 5.0 < LLOQ 995 ± 110 2.89 ± 0.7 132 ± 17 < LLOQ < LLOQ 25.3 ± 1.8 < LLOQ 8.29 ± 0.7 n.s * n.s n.s n.s n.s Wal 80.8 ± 6.4 1.56 ± 0.3 869 ± 78 5.85 ± 1.2 111 ± 14 < LLOQ 2.99 ± 0.5 37.7 ± 4.8 < LLOQ 14.4 ± 2.2 ** ** n.s n.s n.s n.s WalEE 75.8 ± 6.1 1.59 ± 0.3 778 ± 54 7.44 ± 1.4 117 ± 10 < LLOQ 2.91 ± 0.6 22.1 ± 2.9 < LLOQ 7.74 ± 0.8 * *** n.s n.s * n.s B Hydroxy-PUFA [nmol/kg] Epoxy-PUFA [nmol/kg] Liver derived from derived from LA ALA ARA EPA DHA LA ALA  ARA EPA DHA oldCon 114 ± 12 0.65 ± 0.2 150 ± 20 4.47 ± 0.5 28.4 ± 3.9 6.57 ± 0.4 < LLOQ 3.95 ± 0.4 < LLOQ 3.37 ± 0.3 ref ref ref youCon 120 ± 12 0.45 ± 0.1 110 ± 11 4.35 ± 0.6 26.1 ± 3.3 6.40 ± 0.4 < LLOQ 3.37 ± 0.4 < LLOQ 3.13 ± 0.3 n.s n.s n.s n.s n.s n.s n.s n.s Wal 347 ± 47 15.1 ± 2.4 128 ± 13 27.4 ± 2.9 86.0 ± 9.5 12.5 ± 0.8 50.5 ± 10 3.28 ± 0.7 2.15 ± 0.3 13.3 ± 1.3 **** **** n.s **** **** *** n.s **** WalEE 427 ± 50 17.9 ± 1.5 157 ± 17 33.5 ± 3.9 97 ± 8.9 16.5 ± 1.8 54.6 ± 6.7 4.74 ± 0.6 2.45 ± 0.3 17.3 ± 2.0 **** **** n.s **** **** **** n.s **** Significance between groups and the aged control group are displayed as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. LLOQ = lower limit of quantification. Statistical parameters can be found in supplementary Table S4 Cognitive and Motor Function did not improve upon the effects seen in the Wal group, but significantly increased the motor functions in a Rotarod test In the Y-Maze alternation test, oldCon mice were signifi- compared to oldCon. Motor function of Wal mice, however, cantly less active indicated by the number of entries in each was unaffected. Physical activity has already been well arm. They also showed a reduced number of full alterna- established as beneficial to motor and cognitive function tions, reduced spatial memory compared to youCon mice. (Kim et al. 2019; Hatchard et al. 2014; Intlekofer and Cot- Spatial memory was significantly improved in the Wal and man 2013), but to the best of our knowledge a combination WalEE groups. Wang et al. also found that rats adminis- of the two has not yet been reported. Open field experiments tered with 1.1 g/kg, 2.2 g/kg and 11 g/kg walnut oil showed did not show any differences between the groups. Neverthe- improved cognitive performance in Morris Water Maze less these results do not indicate an effect of the interven- (MWM) experiments vaguely based on the concentration tions on mice’s level of curiosity or anxiety/emotionality used (Wang et al. 2018). Other studies in rats and mice also (Seibenhener and Wooten 2015). The mean number of faecal found significant improvements using a 6% walnut-enriched boli of each of the groups reflects this finding. Muthaiyah diet in MWM test (Willis et al. 2009; Muthaiyah et al. 2014) et al., however, reported that mice fed with a 6% walnut- and T-Water maze test (Muthaiyah et al. 2014). Multiple enriched diet showed significantly reduced response to emo- reports have shown that walnuts components, i.e. n3-FA tional adverse events in an elevated-plus-maze experiment (Carey et al. 2013; Eckert et al. 2013), vitamin E (Ashley (Muthaiyah et al. 2014). Looking at our results for the Pas- et al. 2019) or other antioxidative compounds (Ren et al. sive Avoidance test, which applies a small electrical stimulus 2018) like ellagic acid which is metabolized to urolithin A to mice once they cross from a brightly illuminated chamber (Hayes et al. 2016; Gong et al. 2019), showed significant into a dark one, we could clearly see that all groups remem- neuroprotective effects. Although here, walnuts alone did not bered the adverse event on the second day. Unexpectedly, show a significant impact on mice’s motor function, other however, in contrast to an earlier work (Hagl et al. 2016), studies have reported motoric improvements (Willis et al. no significant difference between youCon and oldCon were 2009; Muthaiyah et al. 2014). Increased physical activity observed. As the behaviour of mice differed greatly in each 1 3 NeuroMolecular Medicine (2021) 23:140–160 153 group we propose that the high frequency of tests in a rela- However, the overall effect on gene expression was limited. tively short time (4 different experiments in 4 weeks) might In this study, data suggest that youCon mice tend to show have affected the mice’s behaviour. increased levels of neuronal growth factor, synaptophysin and BDNF as a marker for neuronal growth and synaptic Mitochondrial Function in Mice plasticity. Since young mice are still in development, this difference was to be expected (Reutzel et al. 2020). Concern- Since mitochondrial function plays a key role in neuronal ing the other assessed markers, no distinct effect of Wal or processes during ageing and also the onset of neurodegen- WalEE could be observed. Part of the reason for this might erative diseases (Grimm et al. 2016; Swerdlow et al. 2014), also be the high individual variability of the mice. As these mitochondrial involvement was of key interest. results do not indicate altered cognitive functions, further Multiple studies have shown that fat, the main component behavioural testing should examine and verify the effects of walnuts, and especially n3-FA have a beneficial effect observed in the Y-Maze alternation test. on mitochondrial parameters. Afshordel et al. reported that EPA and DHA from fish oil positively affected the OXPHOS Oxylipin Profiles activity in the brain of aged NMRI mice (Afshordel et al. 2015). Other groups also reported an effect of these fatty Next, we investigated the effect of a walnut-enriched diet on acids on mitochondrial membrane parameters and respira- brain and liver oxylipin profile to get a general understanding tion in human skeletal muscle (Herbst et al. 2014). Ros- of whether there would be a shift to less potent inflamma- signoli et al. also supplemented the diet of mice EPA, DHA tory mediators. Neuroinflammation is a chronic stressor that or conjugated linoleic acid and found beneficial properties occurs during the ageing process (Neves and Sousa-Victor on mitochondrial energetics (Rossignoli et al. 2018). Across 2019; Cole et al. 2009; Hunt et al. 2019; Kim et al. 2016) all mitochondrial parameters investigated in this study, how- and especially n3-PUFA are generally considered to attenu- ever, we found no significant changes and at best only trends ate the inflammatory response by increasing the production for an altered mitochondrial function in aged NMRI mice. of less potent inflammatory mediators or anti-inflammatory Pandaresh et al. who fed tg-AD mice with 6% and a 9% metabolites altogether (Ostermann and Schebb 2017). walnut-enriched diet found that only the higher-dosed diet As it is already well documented in the literature, we affected ROS levels and other markers for oxidative stress are also able to report a significant increase of oxylipins after 5 months of feeding, while the lower dosage which was derived from diet specific PUFA in the intervention groups similar to ours, showed its effect only after 10 and 15 months (Nuernberg et al. 2011; Valencak and Ruf 2011). However, (Pandareesh et al. 2018). As mice in this study were not we found that the oxylipin profile differed in both tissues. transfected and aged normally, mitochondrial dysfunction This is in agreement with Naoe et al. (Naoe et al. 2019), might not be as pronounced as in a dedicated AD mouse who found similar results in liver and brain. In the liver, models. Combined with a feeding time of only 6 months this oxylipins of LA, ALA, EPA and DHA are increased com- might also explain the lack of significant effects. pared to both groups adhering to the control diet, suggest- As ALA is also a major component of walnuts and pre- ing a successful modulation of the liver’s oxylipin profile. cursor to both, EPA and DHA, we were interested to see, Similar, but smaller changes can be seen of the brain. On whether walnuts might attenuated NO-related stress in DBCs one hand walnut-based oxylipins were generally < LLOQ in of aged mice. Nitric oxide radicals induce nitrosative stress control groups while being detectable in both intervention (Carey et al. 2013) and are a crucial part during ageing and groups allowing only a qualitative analysis. On the other in the development of neurodegenerative diseases (Jiménez- hand this confirmed that the brain was affected by the diet as Jiménez et al. 2016). In BV2 microglia cells both EPA and well. Our data suggest that LA, ALA as well as EPA-derived DHA have been found to attenuate NO-related damages (Lu oxylipin levels in the brain were all affected by the interven- et al. 2010; Moon et al. 2007). Introduction of nitrosative tions, while only OH-PUFAs of ARA seemed to be reduced stress via sodium nitroprusside (SNP) showed that neither following walnut feeding and DHA-derived oxylipins were Wal nor WalEE mice protected MMP or ATP level. As the virtually identical with control groups. This is in agreement experimental design did not use a specific part of the brain, with our previous research and a study from Ferdouse et al. but a homogenate of brain cells due to the amounts neces- who also found a general increase in n3-PUFA derived sary for experimentation, we propose that any effects could oxylipins, but little to no changes in ARA-derived oxylipins have been masked by unaffected areas of the brain. (Ostermann et al. 2017a; Ferdouse et al. 2019) in the brain Since the improvements seen in behavioural testing can- of rats. Furthermore, oxylipins derived from DHA, tend to not be explained by an altered mitochondrial function, the be increased only in the liver, but not the brain. However, expression of several genes related to synaptic plasticity, reports have already shown that production of DHA from neuronal growth and antioxidant capacity were investigated. ALA is limited (Demar et al. 2005). This is in line with our 1 3 1 54 NeuroMolecular Medicine (2021) 23:140–160 Fig. 4 Heatmap of hydroxy- PUFA, epoxy-PUFA and dihy- A YouCon oldCon Wal WalEE droxy-PUFA oxylipins derived OH-PUFA 9-HODE 17.9 ± 1.9 19.2 ± 2.6 31.4 ± 2.61 32.2 ± 3.2 from LA, ALA, ARA, EPA and 13-HODE 29.5 ± 3.1 31.0 ± 4.1 49.4 ± 3.8 46.0 ± 4.7 DHA in brain (a) and liver (b). LA Ep-PUFA 9(10)-EpOME 1000% OH-PUFA 4-HDHA 2.35 ± 0.33 1.10 ± 0.15 8.52 ± 1.2 9.69 ± 1.25 250-1000% 7-HDHA 1.09 ± 0.15 1.18 ± 0.145 4.41 ± 0.49 5.11 ± 0.61 110-250% DHA 14-HDHA 4.50 ± 0.86 4.32 ± 0.83 13.2 ± 1.9 13.6 ± 1.3 90-110% 17-HDHA 50% of the samples. One-way ANOVA can be found in supplementary Table S7 with Dunnett’s multiple comparison post-test, comparing all groups in in vivo and in vitro models of AD (Amtul et al. 2012). generally lower effect of the diet on the brain and makes it Therefore, it might be possible that a walnut-enriched diet in difficult to determine the diets effect on the CYP pathway. combination with physical activity might provide a benefit However, looking at the liver increased CYP products could against one of AD’s prominent pathologies. In the brain, be found for LA, ALA as well as several Ep-PUFA of DHA PGD3, which is derived from EPA and formed in the COX and 17(18)-EpETE derived from EPA. Nevertheless, this pathway, is present in small amounts of several mice from modulation had no apparent effect on the ARA-derived CYP both Wal and WalEE groups but was absent in all youCon products in either group or tissue. Of all EpODEs, 15(16)- and oldCon mice. This oxylipin is a less potent inflammatory EpODE was the only oxylipin detectable in both interven- mediator (Ostermann and Schebb 2017) compared to pro- tion groups in the brain and was strongly increased in the inflammatory PGD2 which is formed in the same pathway liver and brain. DiHODEs, the hydrolysis products of EpO- originating from ARA. While Wal mice showed no changes DEs, which can be formed via sEH were increased in the in PGD2 levels, WalEE mice tended to show lower PGD2 same manner. In the liver, both intervention groups showed and significantly lower T xB2 levels. Leng et al. found simi- increased levels of epoxy octadecenoic acids (EpOME) two- lar results and reported decreased PGD2 an T xB2 levels in fold and their dihydroxy-metabolites by two–threefold. The the livers of rats accompanied by increased production of ratio of Ep-PUFA/DiOH-PUFA implies a rather stable sEH P GD3 and TxB3 (Leng et al. 2018). Nevertheless, here, their activity, which further suggest that feeding of walnuts, rich production was not detected in all intervention mice and was in ALA and LA, did not affect sEH in the same way as sup- minimal compared to the high ARA-derived prostanoids. plementation with EPA and DHA did in our previous work Still, their presence in the Wal and WalEE mice hints at (Ostermann and Schebb 2017). a potential modulation towards less potent inflammatory As inflammation is an increasing factor during the age- mediators. ing process, our data show a shift in the oxylipin profile of As Walnuts also contain high amounts of LA, the both brain and liver in line with the dietary supplementa- increases in hydroxy octadecadienoic acids (HODE) were tion. This shift is characterized by increased production of also expected and suggest a modulation of the LOX prod- n3-PUFA oxylipins which are generally considered to be ucts. Our recent work exploring the LOX enzyme affinity beneficial towards inflammation. Nevertheless, the produc- towards DHA supports this shift (Kutzner et al. 2019). As tion of n3-PUFA oxylipins in the brain was smaller com- walnuts were the only source of ALA, its hydroxy-metab- pared to the liver again suggesting that possible effects of olites, i.e. hydroxy octadecatrienoic acids (HOTrE) were the n3-PUFA were very limited and effects on mitochondrial slightly increased in the intervention groups of the brain but function might have been similarly small. Although walnuts more drastically in the liver, while being barely detectable tend to induce promising changes, more pronounced effects in the control groups. might be possible if supplementation of n3-PUFA is directly ALA-based epoxy octadecadienoic acids (EpODEs) were done with EPA and DHA. As we did not look at specific mostly not detectable in the brain, which again supports the regions of the brain to evaluate the oxylipin profiles or the 1 3 1 56 NeuroMolecular Medicine (2021) 23:140–160 Table 6 Ratio of n6/n3-PUFA oxylipins in brain and liver tissue. Dis- the missing change in expression actually has an effect on played are means ± SEM of each groups; n = 14–15. Data were statis- the protein level. Future studies have to show whether the tically compared via an one-way ANOVA with Tukey’s post hoc test and using aged control group (oldCon) as reference (ref.) transcription data also lead to changes at the translation level. Furthermore, Poulose et al. also only found an effect oldCon youCon Wal WalEE in both striatum and hippocampus, when rats were fed with Brain 22.6 ± 1.8 19.7 ± 1.6 20.4 ± 2.8 16.3 ± 1.9 a 9% walnut-enriched diet (Poulose et al. 2013). ref ns ns ns In conclusion, we observed a modulation of the oxylipin Liver 6.29 ± 0.42 5.31 ± 0.33 1.47 ± 0.13 1.79 ± 0.22 profile by a 6% walnut-enriched diet alone and combined ref ns **** **** with physical enrichment in mice. These effects might lead Significance is displayed as: ****p < 0.0001 to better cognitive and motor functions in mice as seen in our oldCon old Control, youCon Young control, Wal Walnut group, results for Rotarod and Y-Maze. Effects of the interventions WalEE Walnut + Enriched Environment group could not be linked to an enhanced mitochondrial function Statistical parameters can be found in supplementary Table S3 and gene expression of markers related to neurogenesis were not affected either. This, however, might be explained with the fact that this study did not investigate specific regions Table 7 Relative normalized mRNA expression of relevant genes from liver tissue. Data are adjusted to aged control group = 100%; of the brain, but a broad selection of different regions in n = 12; Displayed are means ± SEM; one-way ANOVA with which unaffected areas might mask small beneficial effects Dunnett´s multiple comparison post-test comparing all groups with in mitochondria. Furthermore, since results for gene expres- oldCon sion do not indicate altered neuronal function, results from Gene oldCon youCon Wal WalEE Y-maze alternation test should be verified by comparable [%] [%] [%] [%] tests. Results from a recent clinical trial, in which subjects KEAP1 85.3 ± 8.3 118 ± 16 158 ± 25* 138 ± 22 adhered to a walnut-enriched diet for 2 years, did not show NRF2 100 ± 8.7 107 ± 11 77.0 ± 4.6 74.1 ± 5.4 changes in cognition of healthy elderly patients (Sala-Vila CREB1 100 ± 14 133 ± 24 130 ± 21 97.5 ± 11 et al. 2020). Future studies should take an in-depth look IL1β 100 ± 29 52.1 ± 12 56.2 ± 12 47.7 ± 7.0 at different brain regions to isolated specific walnut-related effects. Here, we were able to show that a walnut-enriched Significance is displayed as p < 0.05 *; normalized to the mRNA diet affects cognition in aged NMRI mice and benefits their expression levels of beta-2-microglobulin (B2M) and phosphoglycer- fatty acid composition to potentially attenuate age-related ate kinase 1 (PGK1) neuroinflammation in the ageing brain and body. There- oldCon old Control, youCon Young control, Wal Walnut group, WalEE Walnut + Enriched Environment group fore, walnut-based nutrition might be a promising target for Parameters for the statistical tests can be found in the supplementary healthy ageing, but still needs to be further investigated. Table S8 Acknowledgement We would like to thank the Heinrich Sauer & Josef Schmidt foundation for their support in our research and we would also effect on mitochondrial function, unaffected regions of the like to thank the research group of Zorn et al. (Food Chemistry and brain might mask beneficial effects. Food Biotechnology, JLU Giessen, Germany) for the equipment and know-how provided for food analysis. Looking at the expression of the inflammatory marker IL1β in brain and liver tissue, IL1β levels of either inter- Funding Open Access funding enabled and organized by Projekt vention very virtually identical compared to youCon in DEAL. This research was funded by a grant from the California Wal- the liver. Aged control mice on the other hand showed a nut Commission. greater variability and tended to lean towards elevated IL1β expression. In the brain, the expression of IL1β was virtually Data Availability The dataset generated during this study is available from the corresponding author upon reasonable request. identical across all groups. A recent study by Mejias et al. found that IL1b was increased in aged mice and rats (Mejias Compliance with Ethical Standards et al. 2018; Mawhinney et al. 2011). Lynch et al., as well as Minogue et al. indeed found that EPA reduced hippocampal Conflict of interest The authors declare no conflict of interest. IL1β expression in rats (Lynch et al. 2007; Minogue et al. 2007). Similarly to Moon et al., who reported a reduced Ethical Approval All experiments were carried out by individuals with IL1β expression and PGE2 levels in BV2 microglia upon appropriate training and experience according to the requirements of the Federation of European Laboratory Animal Science Associations EPA treatment, the walnut diet showed similar trends in the and the European Communities Council Directive (Directive 2010/63/ liver for both analytes. Nevertheless, it is uncertain whether EU). Experiments were approved by the regional authority (Regierung- spraesidium Darmstadt; #V54 – 19 c 20/15 – FU/1062). 1 3 NeuroMolecular Medicine (2021) 23:140–160 157 Informed Consent All authors have agreed to the contents of this pub- publication of the Federation of American Societies for Experi- lication. mental Biology, 22(4), 1213–1225. Carey, A. N., Fisher, D. R., Joseph, J. A., & Shukitt-Hale, B. (2013). The ability of walnut extract and fatty acids to protect against the Open Access This article is licensed under a Creative Commons Attri- deleterious effects of oxidative stress and inflammation in hip- bution 4.0 International License, which permits use, sharing, adapta- pocampal cells. Nutritional neuroscience, 16(1), 13–20. tion, distribution and reproduction in any medium or format, as long Cole, G. M., Ma, Q.-L., & Frautschy, S. A. (2009). Omega-3 fatty acids as you give appropriate credit to the original author(s) and the source, and dementia. Prostaglandins, leukotrienes, and essential fatty provide a link to the Creative Commons licence, and indicate if changes acids, 81(2–3), 213–221. were made. The images or other third party material in this article are Coras, R., Murillo-Saich, J. D., & Guma, M. (2020). Circulating Pro- included in the article’s Creative Commons licence, unless indicated and Anti-Inflammatory Metabolites and Its Potential Role in otherwise in a credit line to the material. If material is not included in Rheumatoid Arthritis Pathogenesis. Cells, 9(4), 827. the article’s Creative Commons licence and your intended use is not Costa, D. A., Cracchiolo, J. R., Bachstetter, A. D., Hughes, T. F., Bales, permitted by statutory regulation or exceeds the permitted use, you will K. R., Paul, S. M., et al. (2007). Enrichment improves cognition need to obtain permission directly from the copyright holder. To view a in AD mice by amyloid-related and unrelated mechanisms. Neu- copy of this licence, visit http://creati vecom mons. org/licens es/by/4.0/. robiology of aging, 28(6), 831–844. Cutuli, D. (2017). Functional and Structural Benefits Induced by Omega-3 Polyunsaturated Fatty Acids During Aging. Current neuropharmacology, 15(4), 534–542. Demar, J. C., Ma, K., Chang, L., Bell, J. M., & Rapoport, S. I. (2005). References alpha-Linolenic acid does not contribute appreciably to doco- sahexaenoic acid within brain phospholipids of adult rats fed a Afshordel, S., Hagl, S., Werner, D., Röhner, N., Kögel, D., Bazan, N. diet enriched in docosahexaenoic acid. Journal of neurochemis- G., et al. (2015). Omega-3 polyunsaturated fatty acids improve try, 94(4), 1063–1076. mitochondrial dysfunction in brain aging–impact of Bcl-2 and Desai, A., Park, T., Barnes, J., Kevala, K., Chen, H., & Kim, H.-Y. NPD-1 like metabolites. Prostaglandins, leukotrienes, and essen- (2016). Reduced acute neuroinflammation and improved func- tial fatty acids, 92, 23–31. tional recovery after traumatic brain injury by α-linolenic acid Ali, W., Ikram, M., Park, H. Y., Jo, M. G., Ullah, R., Ahmad, S., et al. supplementation in mice. Journal of neuroinflammation, 13(1), (2020). Oral Administration of Alpha Linoleic Acid Rescues 253. Aβ-Induced Glia-Mediated Neuroinflammation and Cognitive Dilberger, B., Passon, M., Asseburg, H., Silaidos, C. V., Schmitt, Dysfunction in C57BL/6N Mice. Cells, 9(3), 667. F., Schmiedl, T., et al. (2019). Polyphenols and Metabolites Amtul, Z., Uhrig, M., Wang, L., Rozmahel, R. F., & Beyreuther, K. Enhance Survival in Rodents and Nematodes-Impact of Mito- (2012). Detrimental effects of arachidonic acid and its metabo- chondria. Nutrients, 11(8), 1886. lites in cellular and mouse models of Alzheimer’s disease: struc- Dilberger, B., Baumanns, S., Schmitt, F., Schmiedl, T., Hardt, M., tural insight. Neurobiology of aging, 33(4), 831.e21–31. Wenzel, U., et  al. (2019). Mitochondrial Oxidative Stress Ashley, S., Bradburn, S., & Murgatroyd, C. (2019). A meta-analysis Impairs Energy Metabolism and Reduces Stress Resistance of peripheral tocopherol levels in age-related cognitive decline and Longevity of C. elegans. Oxidative medicine and cellular and Alzheimer’s disease. Nutritional neuroscience. https ://doi. longevity, 2019, 6840540. org/10.1080/102841 5X.2019.16810 66. Dyall, S. C. (2015). Long-chain omega-3 fatty acids and the brain: a Asseburg, H., Schäfer, C., Müller, M., Hagl, S., Pohland, M., Berres- review of the independent and shared effects of EPA, DPA and sem, D., et al. (2016). Effects of Grape Skin Extract on Age- DHA. Frontiers in aging neuroscience, 7, 52. Related Mitochondrial Dysfunction, Memory and Life Span in Eckert, G. P., Franke, C., Nöldner, M., Rau, O., Wurglics, M., Schu- C57BL/6J Mice. Neuromolecular medicine, 18(3), 378–395. bert-Zsilavecz, M., et al. (2010). Plant derived omega-3-fatty Baraldi, T., Schöwe, N. M., Balthazar, J., Monteiro-Silva, K. C., Albu- acids protect mitochondrial function in the brain. Pharmaco- querque, M. S., Buck, H. S., et al. (2013). Cognitive stimulation logical research, 61(3), 234–241. during lifetime and in the aged phase improved spatial memory, Eckert, G. P., Chang, S., Eckmann, J., Copanaki, E., Hagl, S., Hener, and altered neuroplasticity and cholinergic markers of mice. U., et al. (2011). Liposome-incorporated DHA increases neu- Experimental gerontology, 48(8), 831–838. ronal survival by enhancing non-amyloidogenic APP process- Blondeau, N., Nguemeni, C., Debruyne, D. N., Piens, M., Wu, X., ing. Biochimica et biophysica acta, 1808(1), 236–243. Pan, H., et al. (2009). Subchronic alpha-linolenic acid treatment Eckert, G. P., Lipka, U., & Muller, W. E. (2013). Omega-3 fatty enhances brain plasticity and exerts an antidepressant effect: a acids in neurodegenerative diseases: focus on mitochondria. versatile potential therapy for stroke. Neuropsychopharmacology Prostaglandins, leukotrienes, and essential fatty acids, 88(1), official publication of the American College of Neuropsychop- 105–114. harmacology, 34(12), 2548–2559. Emendato, A., Spadaccini, R., de Santis, A., Guerrini, R., D’Errico, Bourre, J. M. (2006a). Effects of nutrients (in food) on the structure and G., & Picone, D. (2016). Preferential interaction of the Alzhei- function of the nervous system: update on dietary requirements mer peptide Aβ-(1–42) with Omega-3-containing lipid bilayers: for brain. Part 1: micronutrients. The journal of nutrition, health structure and interaction studies. FEBS letters, 590(4), 582–591. & aging, 10(5), 377–85. Ferdouse, A., Leng, S., Winter, T., & Aukema, H. M. (2019). The Bourre, J. M. (2006b). Effects of nutrients (in food) on the structure and Brain Oxylipin Profile Is Resistant to Modulation by Dietary n-6 function of the nervous system: update on dietary requirements and n-3 Polyunsaturated Fatty Acids in Male and Female Rats. for brain. Part 2 macronutrients. The journal of nutrition, health Lipids, 54(1), 67–80. & aging, 10(5), 386–99. Franke, C., Nöldner, M., Abdel-Kader, R., Johnson-Anuna, L. N., Bousquet, M., Saint-Pierre, M., Julien, C., Salem, N., Cicchetti, F., & Gibson Wood, W., Müller, W. E., et al. (2007). Bcl-2 upregula- Calon, F. (2008). Beneficial effects of dietary omega-3 polyun- tion and neuroprotection in guinea pig brain following chronic saturated fatty acid on toxin-induced neuronal degeneration in simvastatin treatment. Neurobiology of disease, 25(2), 438–445. an animal model of Parkinson’s disease. FASEB journal official 1 3 158 NeuroMolecular Medicine (2021) 23:140–160 Gabbs, M., Leng, S., Devassy, J. G., Monirujjaman, M., & Aukema, H. Mediators in Biological Samples. Frontiers in Pharmacology. M. (2015). Advances in Our Understanding of Oxylipins Derived https ://doi.org/10.3389/fphar. 2019.00169. from Dietary PUFAs. Advances in nutrition (Bethesda, Md.), Leng, S., Winter, T., & Aukema, H. M. (2017). Dietary LA and sex 6(5), 513–540. effects on oxylipin profiles in rat kidney, liver, and serum differ Gnaiger, E. (2020). Mitochondrial pathways and respiratory control An from their effects on PUFAs. Journal of lipid research, 58(8), introduction to OXPHOS analysis Bioenerg Commun. Innsbruck: 1702–1712. OROBOROS MiPNet Publications. Leng, S., Winter, T., & Aukema, H. M. (2018). Dietary ALA, EPA Gong, Z., Huang, J., Xu, B., Ou, Z., Zhang, L., Lin, X., et al. (2019). and DHA have distinct effects on oxylipin profiles in female and Urolithin A attenuates memory impairment and neuroinflamma- male rat kidney, liver and serum. The Journal of nutritional bio- tion in APP mice. Journal of neuroinflammation, 16(1), 62. chemistry, 57, 228–237. Grimm, A., Friedland, K., & Eckert, A. (2016). Mitochondrial dysfunc- Litwiniuk, A., Domańska, A., Chmielowska, M., Martyńska, L., Bik, tion: the missing link between aging and sporadic Alzheimer’s W., & Kalisz, M. (2020). The Effects of Alpha-Linolenic Acid disease. Biogerontology, 17(2), 281–296. on the Secretory Activity of Astrocytes and β Amyloid-Asso- Grimm, A., Friedland, K., & Eckert, A. (2016). Mitochondrial dysfunc- ciated Neurodegeneration in Differentiated SH-SY5Y Cells: tion: the missing link between aging and sporadic Alzheimer’s Alpha-Linolenic Acid Protects the SH-SY5Y cells against β disease. Biogerontology, 17(2), 281–96. Amyloid Toxicity. Oxidative medicine and cellular longevity, Hagl, S., Asseburg, H., Heinrich, M., Sus, N., Blumrich, E.-M., Drin- 2020, 8908901. gen, R., et al. (2016). Effects of Long-Term Rice Bran Extract Liu, J., Chen, D., Wang, Z., Chen, C., Ning, D., & Zhao, S. (2019). Supplementation on Survival, Cognition and Brain Mitochon- Protective effect of walnut on d-galactose-induced aging mouse drial Function in Aged NMRI Mice. Neuromolecular medicine, model. Food Science & Nutrition, 7(3), 969–976. 18(3), 347–363. Lores-Arnaiz, S., Lores Arnaiz, M. R., Czerniczyniec, A., Cuello, M., Haider, S., Batool, Z., Tabassum, S., Perveen, T., Saleem, S., Naqvi, F., & Bustamante, J. (2010). Mitochondrial function and nitric oxide et al. (2011). Effects of walnuts (Juglans regia) on learning and production in hippocampus and cerebral cortex of rats exposed to memory functions. Plant foods for human nutrition (Dordrecht, enriched environment. Brain research, 1319, 44–53. Netherlands), 66(4), 335–340. Lu, D.-Y., Tsao, Y.-Y., Leung, Y.-M., & Su, K.-P. (2010). Docosahex- Hatchard, T., Ting, J. J., & Messier, C. (2014). Translating the impact aenoic acid suppresses neuroinflammatory responses and induces of exercise on cognition: Methodological issues in animal heme oxygenase-1 expression in BV-2 microglia: implications of research. Behavioural Brain Research, 273, 177–88. antidepressant effects for ω-3 fatty acids. Neuropsychopharma- Hayes, D., Angove, M. J., Tucci, J., & Dennis, C. (2016). Walnuts cology official publication of the American College of Neuropsy- (Juglans regia) Chemical Composition and Research in Human chopharmacology, 35(11), 2238–2248. Health. Critical reviews in food science and nutrition, 56(8), Lynch, A. M., Loane, D. J., Minogue, A. M., Clarke, R. M., Kilroy, D., 1231–1241. Nally, R. E., et al. (2007). Eicosapentaenoic acid confers neuro- Herbst, E. A. F., Paglialunga, S., Gerling, C., Whitfield, J., Mukai, K., protection in the amyloid-beta challenged aged hippocampus. Chabowski, A., et al. (2014). Omega-3 supplementation alters Neurobiology of aging, 28(6), 845–855. mitochondrial membrane composition and respiration kinetics Maesako, M., Uemura, K., Kubota, M., Kuzuya, A., Sasaki, K., Asada, in human skeletal muscle. The Journal of physiology, 592(6), M., et al. (2012). Environmental enrichment ameliorated high-fat 1341–1352. diet-induced Aβ deposition and memory deficit in APP trans- Hunt, N. J., Kang, S. W. S., Lockwood, G. P., Le Couteur, D. G., & genic mice. Neurobiology of aging, 33(5), 1011.e11–23. Cogger, V. C. (2019). Hallmarks of Aging in the Liver. Compu- Martínez-Lapiscina, E. H., Clavero, P., Toledo, E., Estruch, R., Salas- tational and structural biotechnology journal, 17, 1151–1161. Salvadó, J., San Julián, B., et al. (2013). Mediterranean diet Intlekofer, K. A., & Cotman, C. W. (2013). Exercise counteracts declin- improves cognition: the PREDIMED-NAVARRA randomised ing hippocampal function in aging and Alzheimer’s disease. Neu- trial. Journal of Neurology, Neurosurgery and Psychiatry, robiology of Diseases, 57, 47–55. 84(12), 1318–1325. Jiménez-Jiménez, F. J., Alonso-Navarro, H., Herrero, M. T., García- Mawhinney, L. J., de Vaccari, J. P. R., Dale, G. A., Keane, R. W., & Martín, E., & Agúndez, J. A. G. (2016). An Update on the Role Bramlett, H. M. (2011). Heightened inflammasome activation is of Nitric Oxide in the Neurodegenerative Processes of Parkin- linked to age-related cognitive impairment in Fischer 344 rats. son’s Disease. Current medicinal chemistry, 23(24), 2666–2679. BMC neuroscience, 12, 123. Kim, I. H., Xu, J., Liu, X., Koyama, Y., Ma, H.-Y., Diggle, K., et al. McKay, D. L., Chen, C.-Y.O., Yeum, K.-J., Matthan, N. R., Lichten- (2016). Aging increases the susceptibility of hepatic inflamma- stein, A. H., & Blumberg, J. B. (2010). Chronic and acute effects tion, liver fibrosis and aging in response to high-fat diet in mice. of walnuts on antioxidant capacity and nutritional status in Age (Dordrecht, Netherlands), 38(4), 291–302. humans a randomized, crossover pilot study. Nutrition journal, Kim, D., Cho, J., & Kang, H. (2019). Protective effect of exercise train- 9, 21. ing against the progression of Alzheimer’s disease in 3xTg-AD Mejias, N. H., Martinez, C. C., Stephens, M. E., Rivero Vaccari, J. P., & mice. Behavioural Brain Research, 374, 112105. de. . (2018). Contribution of the inflammasome to inflammaging. Koch, E., Mainka, M., Dalle, C., Ostermann, A. I., Rund, K. M., Journal of inflammation (London, England), 15, 23. Kutzner, L., et al. (2020). Stability of oxylipins during plasma Mendonça, A. M., Cayer, L. G. J., Pauls, S. D., Winter, T., Leng, S., generation and long-term storage. Talanta, 217, 121074. Taylor, C. G., et al. (2018). Distinct effects of dietary ALA, EPA Kutzner, L., Goloshchapova, K., Heydeck, D., Stehling, S., Kuhn, H., and DHA on rat adipose oxylipins vary by depot location and & Schebb, N. H. (2017). Mammalian ALOX orthologs exhibit sex. Prostaglandins, leukotrienes, and essential fatty acids, 129, pronounced dual positional specificity with docosahexaenoic 13–24. acid. Biochimica et Biophysica Acta (BBA)–Molecular and Cell Minciullo, P. L., Catalano, A., Mandraffino, G., Casciaro, M., Crucitti, Biology of Lipids, 1862(7), 666–675. A., Maltese, G., et al. (2016). Inflammaging and Anti-Inflam- Kutzner, L., Rund, K. M., Ostermann, A. I., Hartung, N. M., Galano, maging: The Role of Cytokines in Extreme Longevity. Archivum J.-M., Balas, L., et al. (2019). Development of an Optimized immunologiae et therapiae experimentalis, 64(2), 111–126. LC-MS Method for the Detection of Specialized Pro-Resolving 1 3 NeuroMolecular Medicine (2021) 23:140–160 159 Minogue, A. M., Lynch, A. M., Loane, D. J., Herron, C. E., & Lynch, Reutzel, M., Grewal, R., Dilberger, B., Silaidos, C., Joppe, A., & Eck- M. A. (2007). Modulation of amyloid-beta-induced and age- ert, G. P. (2020). Cerebral Mitochondrial Function and Cogni- associated changes in rat hippocampus by eicosapentaenoic acid. tive Performance during Aging: A Longitudinal Study in NMRI Journal of neurochemistry, 103(3), 914–926. Mice. Oxidative medicine and cellular longevity, 2020, 4060769. Moon, D.-O., Kim, K.-C., Jin, C.-Y., Han, M.-H., Park, C., Lee, K.-J., Rey, C., Delpech, J. C., Madore, C., Nadjar, A., Greenhalgh, A. D., et al. (2007). Inhibitory effects of eicosapentaenoic acid on Amadieu, C., et al. (2019). Dietary n-3 long chain PUFA sup- lipopolysaccharide-induced activation in BV2 microglia. Inter- plementation promotes a pro-resolving oxylipin profile in the national immunopharmacology, 7(2), 222–229. brain. Brain, behavior, and immunity, 76, 17–27. Muthaiyah, B., Essa, M. M., Lee, M., Chauhan, V., Kaur, K., & Chau- Ros, E. (2009). Nuts and novel biomarkers of cardiovascular disease. han, A. (2014). Dietary supplementation of walnuts improves The American journal of clinical nutrition, 89(5), 1649S-S1656. memory deficits and learning skills in transgenic mouse model of Rosales-Martínez, P., Arellano-Cárdenas, S., Dorantes-Álvarez, L., Alzheimer’s disease. Journal of Alzheimer’s disease JAD, 42(4), García-Ochoa, F., & López-Cortez, M. D. S. (2017). Compari- 1397–1405. son Between Antioxidant Activities of Phenolic Extracts from Naoe, S., Tsugawa, H., Takahashi, M., Ikeda, K., & Arita, M. (2019). Mexican Peanuts, Peanuts Skins, Nuts and Pistachios. Journal Characterization of Lipid Profiles after Dietary Intake of Polyun- of the Mexican Chemical Society, 58(2), 185–193. saturated Fatty Acids Using Integrated Untargeted and Targeted Rossignoli, C. P., Dechandt, C. R. P., Souza, A. O., Sampaio, I. H., Lipidomics. Metabolites, 9(10), 241. Vicentini, T. M., Teodoro, B. G., et al. (2018). Effects of inter- Nayeem, M. A. (2018). Role of oxylipins in cardiovascular diseases. mittent dietary supplementation with conjugated linoleic acid Acta pharmacologica Sinica, 39(7), 1142–1154. and fish oil (EPA/DHA) on body metabolism and mitochondrial Neves, J., & Sousa-Victor, P. (2019). Regulation of inflammation as energetics in mice. The Journal of nutritional biochemistry, 60, an anti-aging intervention. The FEBS journal, 287(1), 43–52. 16–23. Nuernberg, K., Breier, B. H., Jayasinghe, S. N., Bergmann, H., Thomp- Rund, K. M., Ostermann, A. I., Kutzner, L., Galano, J.-M., Oger, C., son, N., Nuernberg, G., et al. (2011). Metabolic responses to Vigor, C., et al. (2018). Development of an LC-ESI(-)-MS/MS high-fat diets rich in n-3 or n-6 long-chain polyunsaturated fatty method for the simultaneous quantification of 35 isoprostanes acids in mice selected for either high body weight or leanness and isofurans derived from the major n3- and n6-PUFAs. Ana- explain different health outcomes. Nutrition & metabolism, 8(1), lytica chimica acta, 1037, 63–74. 56. Rund, K. M., Heylmann, D., Seiwert, N., Wecklein, S., Oger, C., Official Methods of Analysis of AOAC International. AOAC Official Galano, J.-M., et al. (2019). Formation of trans-epoxy fatty Method 963.15 -Fat in Cacao Products 1995. acids correlates with formation of isoprostanes and could serve Ostermann, A. I., & Schebb, N. H. (2017). Effects of omega-3 fatty as biomarker of oxidative stress. Prostaglandins & Other Lipid acid supplementation on the pattern of oxylipins: a short review Mediators, 144, 106334. about the modulation of hydroxy-, dihydroxy-, and epoxy-fatty Sala-Vila, A., Valls-Pedret, C., Rajaram, S., Coll-Padrós, N., Cofán, acids. Food & function, 8(7), 2355–2367. M., Serra-Mir, M., et al. (2020). Effect of a 2-year diet inter- Ostermann, A. I., Waindok, P., Schmidt, M. J., Chiu, C.-Y., Smyl, vention with walnuts on cognitive decline. The Walnuts And C., Rohwer, N., et al. (2017a). Modulation of the endogenous Healthy Aging (WAHA) study: a randomized controlled trial. omega-3 fatty acid and oxylipin profile in vivo-A comparison of The American journal of clinical nutrition, 111(3), 590–600. the fat-1 transgenic mouse with C57BL/6 wildtype mice on an Sales-Campos, H., deSouza, P. R., Peghini, B. C., da Silva, J. S., & omega-3 fatty acid enriched diet. PLoS ONE, 12(9), e0184470. Cardoso, C. R. (2013). An Overview of the Modulatory Effects Ostermann, A. I., Reutzel, M., Hartung, N., Franke, N., Kutzner, L., of Oleic Acid in Health and Disease. MRMC, 13(2), 201–210. Schoenfeld, K., et al. (2017b). A diet rich in omega-3 fatty acids Samson, A. L., Ju, L., Ah Kim, H., Zhang, S. R., Lee, J. A. A., Stur- enhances expression of soluble epoxide hydrolase in murine geon, S. A., et al. (2015). MouseMove: an open source program brain. Prostaglandins & Other Lipid Mediators, 133, 79–87. for semi-automated analysis of movement and cognitive testing Ostermann, A. I., Koch, E., Rund, K. M., Kutzner, L., Mainka, M., & in rodents. Scientific reports, 5, 16171. Schebb, N. H. (2020). Targeting esterified oxylipins by LC–MS - Schaffer, S., Asseburg, H., Kuntz, S., Muller, W. E., & Eckert, G. Effect of sample preparation on oxylipin pattern. Prostaglandins P. (2012). Effects of polyphenols on brain ageing and Alzhei- & Other Lipid Mediators, 146, 106384. mer’s disease: focus on mitochondria. Molecular neurobiology, Pandareesh, M. D., Chauhan, V., & Chauhan, A. (2018). Walnut 46(1), 161–178. Supplementation in the Diet Reduces Oxidative Damage and Seibenhener, M. L., & Wooten, M. C. (2015). Use of the Open Field Improves Antioxidant Status in Transgenic Mouse Model of Maze to measure locomotor and anxiety-like behavior in mice. Alzheimer’s Disease. Journal of Alzheimer’s disease JAD, 64(4), Journal of visualized experiments JoVE, 96, e52434. 1295–1305. Shashikumar, S., Pradeep, H., Chinnu, S., Rajini, P. S., & Rajanikant, Poulose, S. M., Bielinski, D. F., & Shukitt-Hale, B. (2013). Walnut diet G. K. (2015). Alpha-linolenic acid suppresses dopaminergic reduces accumulation of polyubiquitinated proteins and inflam- neurodegeneration induced by 6-OHDA in C. elegans. Physiol- mation in the brain of aged rats. The Journal of nutritional bio- ogy & Behavior, 151, 563–569. chemistry, 24(5), 912–919. Signini, É. F., Nieman, D. C., Silva, C. D., Sakaguchi, C. A., & Poulose, S. M., Miller, M. G., & Shukitt-Hale, B. (2014). Role of wal- Catai, A. M. (2020). Oxylipin Response to Acute and Chronic nuts in maintaining brain health with age. The Journal of nutri- Exercise: A Systematic Review. Metabolites, 10(6), 264. tion, 144(4 Suppl), 561S-566S. Smarr, B., Rowland, N. E., & Zucker, I. (2019). Male and female Prendergast, B. J., Onishi, K. G., & Zucker, I. (2014). Female mice mice show equal variability in food intake across 4-day spans liberated for inclusion in neuroscience and biomedical research. that encompass estrous cycles. PLoS ONE, 14(7), e0218935. Neuroscience and biobehavioral reviews, 40, 1–5. Steiner, J. L., Murphy, E. A., McClellan, J. L., Carmichael, M. D., Ren, D., Zhao, F., Liu, C., Wang, J., Guo, Y., Liu, J., et al. (2018). Anti- & Davis, J. M. (2011). Exercise training increases mitochon- oxidant hydrolyzed peptides from Manchurian walnut (Juglans drial biogenesis in the brain. Journal of applied physiology mandshurica Maxim.) attenuate scopolamine-induced memory (Bethesda, Md. 1985), 111(4), 1066–71. impairment in mice. Journal of the science of food and agricul- Swerdlow, R. H., Burns, J. M., & Khan, S. M. (2014). The Alz- ture, 98(13), 5142–52. heimer’s disease mitochondrial cascade hypothesis: progress 1 3 1 60 NeuroMolecular Medicine (2021) 23:140–160 and perspectives. Biochimica et biophysica acta, 1842(8), Verret, L., Krezymon, A., Halley, H., Trouche, S., Zerwas, M., 1219–1231. Lazouret, M., et al. (2013). Transient enriched housing before Thomas, J., Thomas, C. J., Radcliffe, J., & Itsiopoulos, C. (2015). amyloidosis onset sustains cognitive improvement in Tg2576 Omega-3 Fatty Acids in Early Prevention of Inflammatory mice. Neurobiology of aging, 34(1), 211–225. Neurodegenerative Disease: A Focus on Alzheimer’s Disease. Vinson, J. A., & Cai, Y. (2012). Nuts, especially walnuts, have both BioMed research international, 2015, 172801. antioxidant quantity and efficacy and exhibit significant potential Valencak, T. G., & Ruf, T. (2011). Feeding into old age: long-term health benefits. Food & function, 3(2), 134–140. effects of dietary fatty acid supplementation on tissue composi- Wang, L.-M., Yi, Y., Yao, Y.-L., Feng, G., Shu, C., Wang, H.-X., et al. tion and life span in mice. Journal of comparative physiology. (2018). Walnut oil improves spatial memory in rats and increases B, Biochemical, systemic, and environmental physiology, 181(2), the expression of acid-sensing ion channel genes Asic2a and 289–98. Asic4. Food Science & Nutrition, 7(1), 293–301. Valls-Pedret, C., Lamuela-Raventós, R. M., Medina-Remón, A., Quin- Weiser, M. J., Butt, C. M., & Mohajeri, M. H. (2016). Docosahexaenoic tana, M., Corella, D., Pintó, X., et al. (2012). Polyphenol-rich Acid and Cognition throughout the Lifespan. Nutrients, 8(2), 99. foods in the Mediterranean diet are associated with better cog- Willis, L. M., Shukitt-Hale, B., Cheng, V., & Joseph, J. A. (2009). nitive function in elderly subjects at high cardiovascular risk. Dose-dependent effects of walnuts on motor and cognitive Journal of Alzheimer’s disease JAD, 29(4), 773–782. function in aged rats. The British journal of nutrition, 101(8), Valls-Pedret, C., Sala-Vila, A., Serra-Mir, M., Corella, D., La Torre, 1140–1144. R., de, Martínez-González MÁ, , et al. (2015). Mediterranean Diet and Age-Related Cognitive Decline: A Randomized Clinical Publisher’s Note Springer Nature remains neutral with regard to Trial. JAMA internal medicine, 175(7), 1094–1103. jurisdictional claims in published maps and institutional affiliations. 1 3