Effects of Dietary Supplementation with EPA-enriched Phosphatidylcholine and Phosphatidylethanolamine on Glycerophospholipid Profile in Cerebral Cortex of SAMP8 Mice fed with High-fat Diet

: The destruction of lipid homeostasis is associated with nervous system diseases such as Alzheimer’s disease (AD). It has been reported that dietary EPA-enriched phosphatidylcholine (EPA-PC) and phosphatidylethanolamine (EPA-PE) could improve brain function. However, it was unclear that whether EPA-PC and EPA-PE intervention could change the lipid composition of cerebral cortex in AD mice. All the senescence-accelerated mouse-prone 8 (SAMP8) mice were fed with a high-fat diet for 8 weeks. After another 8 weeks of intervention with EPA-PC and EPA-PE (1%, w/w), the cerebral cortex lipid levels were determined by lipidomics. Results demonstrated that dietary supplementation with EPA-PE and EPA-PC for 8 weeks significantly increased the amount of choline plasmalogen (pPC) and Lyso-phosphatidylethanolamine (LPE) in the cerebral cortex of SAMP8 mice fed with high fat diet. Meanwhile, administration with EPA-PE and EPA-PC could significantly decrease the level of docosapentaenoic acid (DPA)-containing phosphatidylserine (PS) as well as increase the levels of arachidonic acid (AA)-containing phosphatidylethanolamine and PS in cerebral cortex. EPA-PE and EPA-PC could restore the lipid homeostasis of dementia mice to a certain degree, which might provide a potential novel therapy strategy and direction of dietary intervention in patients with cognitive impairment.


Introduction
Alzheimer s disease AD , a neurodegenerative disease, has the main manifestations including progressive memory loss, cognitive dysfunction, inattention, emotional disorders, personality changes and other characteristics 1 . As AD patients cannot take care of themselves, especially in the late stage of disease, long-term care of Alzheimer s patients will be a serious social burden in the aging society. It has been reported that dietary intervention can affect the occurrence and development of AD 2 . Saturated fatty acid intake may increase the AD risk, while intake of unsaturated fatty acids may play an important role in alleviating AD 3 . Some studies have shown that senescence-accelerated mouse-prone 8 SAMP8 mice fed with high fat diet exhibited higher cerebral amyloid beta Aβ level, dysregulated tau phosphorylating glycogen synthase kinase 3β and reduced synaptophysin immunoreactivity, as well as showed the serious learning and memory disorders compared to senescence-accelerated mouse-prone 8 SAMP8 mice fed with low fat diet 4,5 . In addition, high fat diet could significantly regulate the content of lipid subclasses in the brain of Alzheimer s model rats 6 .
Aβ plaques are the most characteristic cellular changes observed in Alzheimer s disease. Aβ is produced as a secretory product of amyloid precursor protein APP by sequential cleavage under the action of α-, β-, and γ-secretase embedded in the neuronal cell membrane. Studies have demonstrated that the production and uptake of Aβ were highly associated with the composition and biophysical properties of the membrane. Fatty acids with four or more double bonds, such as arachidonic acid AA , eicosapentaenoic acid EPA , and docosahexaenoic acid DHA , could increase membrane fluidity and lead to an increase of neuroprotective soluble APP secretion 7 . Studies have shown that dietary supplementation with EPA and DHA could reduce the risk of AD 8 . However, most studies focused on the protective effect of DHA on AD, while the protective effect of EPA was rarely involved 9 . Compared with EPAenriched ethyl esters EPA-EE and EPA-enriched triacylglycerol EPA-TG , EPA-enriched phospholipids EPA-PL could alleviate Aβ-induced cognitive impairment 9 . Our previous studies have shown that EPA-enriched phosphatidylcholine EPA-PC and EPA-enriched phosphatidylethanolamine EPA-PE could improve the memory and cognitive function of SAMP8 mice fed with high fat diet by inhibiting Aβ production, oxidative stress and apoptosis, downregulating inflammatory reaction and increasing neuronal activity 10 12 . Naturally, neurological disorders are associated with lipid signaling, metabolism, trafficking, and homeostasis. However, there were no reports about the effects of dietary supplementation with EPA-PC and EPA-PE on lipid profiles in the brain of dementia animals.
Lipidomics could be used to determine lipid profiles in pathological process and might demonstrate the functional properties in different physiological status. SAMP8 mouse is a well-known animal model of AD, which has similarities with AD patients in many aspects 13 . The purpose of this study was to investigate the effects of dietary EPA-PC and EPA-PE on the AD brain lipid profiles, and the lipid composition in cerebral cortex of high-fat diet-induced SAMP8 mice with cognitive deficiency was evaluated by lipidomic analysis.
2 Materials and Method 2.1 EPA-PC and EPA-PE preparation EPA-enriched phospholipids were isolated from sea cucumber Cucumaria frondosa Nanshan Aquatic Market, Qingdao, China . Total lipids was extracted from the comminuted sea cucumber by chloroform-methanol 2:1 for 24 hours according to the previous method 14 . Then the extraction was mixed with NaCl solution 0.15M and kept for 24 h to obtain the chloroform phase. EPA-PC was purified from total lipids by silica-gel column chromatography using chloroform, chloroform-methanol 9:1 , acetone, chloroform-methanol 2:1 and methanol sequentially as eluents. EPA-PC was obtained after removal of the methanol eluent under vacuum, respectively. EPA-PE was enzymatically synthesized by catalyzing transphosphatidylation of EPA-PC using phospholipase D in ethanolamine-containing buffer system. The purity of EPA-PC and EPA-PE was above 90 analyzed by high-performance liquid chromatography with evaporative light scattering detector HPLC-ELSD . The fatty acid composition of EPA-PC and EPA-PE was determined by an Agilent 6890 gas chromatography with a flame-ionization detector and the relative content of each fatty acid was shown in Table 1.

Animals and diet
Male SAMP8 mice purchased from NanjingQingzilan Technology Co., Ltd were housed at SPF environment under the temperature of 20 2 and the humidity of 60 with a 12/12 h light/dark cycle light starting at 8 a.m. . SAMP8 mice were fed with high fat diet for 8 weeks, and then were randomly divided into three groups including high fat group HF , EPA-PC group and EPA-PE group n 8/group . The diet of the three groups was prepared according to the AIN-93M rodent feed formula. Mice in EPA-PC group and EPA-PE group were fed with high fat diet containing 1 w/w EPA-PC and 1 w/w EPA-PE, respectively. The mice were continuously treated with corresponding diets for 8 weeks. The study protocols were approved by the ethical committee of experimental animal care at Ocean University of China Qingdao, China, approval no. SPXY2019031 . The ingredients and fatty acid composition of experimental diets are shown in Tables 2 and 3, respectively. Note: "-" , not detected. Note: "-" , not detected.

Tissue sample preparation and lipidomic analysis
After eight weeks of intervention, the mice were anesthetized by an intraperitoneal injection of sodium chloride solution containing 1 pentobarbital sodium, and then all the mice were decapitated. Cortex was quickly separated from the brain. Cortex was crushed in phosphate buffer using tissue homogenizer and then was extracted with chloroform:methanol 2:1, v/v . NP-HPLC was used to separate the polar lipids from total lipids using Phenomenex Luna 3 μm micron silica column 150 2.0 mm . Mobile phase A was chloroform:methanol:ammonia water at the ratio of 89.5:10:0.5, and mobile phase B was chloroform: ethanol:ammonia water:water at the ratio of 55:39:0.5:5.5. The flow rate was 0.25 mL/min. The gradient of mobile phase A was maintained at 95 for 5 min, then decreased linearly to 60 within 7 min and kept for 4 min. The gradient of mobile phase A was further increased to 70 and kept for 15 minutes. Finally, re-equilibration was performed from 33 min to 38 min with 95 A. All experiments were carried out by Exion UPLC-QTRAP 6500 PLUS Sciex liquid chromatography-mass spectrometry, and the analyses were performed in the electrospray ionization ESI mode. The conditions were: curtain gas 20 psi, ion spray voltage 5500 V, temperature 400 , ion source gas 35 psi, ion source gas 2 35 psi 15 .

Statistical analysis
Statistical analyses were performed using MATLAB R2019b MathWorks, USA . The differences between groups were tested by one-way ANOVA Turkey s test . P 0.05 was considered to indicate a statistically significant difference.

Alterations of glycerophospholipids
There was no significant difference in body weight, food intake and brain weight among the HF group, EPA-PC group and EPA-PE group Table 4 . Brain is the most lipidrich organ except for the adipose tissue, and brain lipids account for at least 50 of the dry brain weight. Brain lipids are comprised of 50 phospholipids, 40 glycolipids, 10 cholesterol, cholesterol ester and traces of triacylglycerols 6 . Glycerophospholipid, such as phosphatidyl-choline PC and phosphatidylethanolamine PE , is one of the most important lipids in cerebral cortex. Destruction of lipid homeostasis is related to neurodegenerative diseases, such as Alzheimer s disease 16 . It is of vital importance to illustrate the changes of lipid profile in brain after dietary intervention.
In order to illustrate the effects of EPA-PC and EPA-PE on AD brain lipid profiles in SAMP8 mice fed with high-fat diet, the lipid composition in cerebral cortex was detected by lipidomic analysis. We mainly identified 9 classes of glycerophospholipids comprised of 170 molecular species. After 8 weeks of EPA-PC and EPA-PE intervention, the changes of brain glycerophospholipid in AD model mice were shown in Fig. 1. There were only small changes without significance in the levels of the main glycerophospholipid classes. PC and PE, accounting for more than 60 of the total glycerophospholipid, were the primary ingredients in the cerebral cortex, which were not easily affected by exogenous phospholipid intake  EPA could not affect the molecular species of brain phospholipids, especially PC, phosphatidylserine PS , PE and lyso-phosphatidylcholine LPC 17 , which was consistent with the present study. Notably, dietary supplementation with EPA-PE and EPA-PC for 8 weeks significantly increased the amount of choline plasmalogen pPC and lysophosphatidylethanolamine LPE in the cerebral cortex although the content was relatively low. Several studies showed a decrease of pPC content in AD brains 18 , and the increase of pPC content after dietary intervention suggested the therapeutic effect of EPA-PE and EPA-PC on AD.  Table 5. A metabonomics study found that patients with AD and mild cognitive impairment MCI had lower PC 16:0/22:6 and PC 18:0/22:6 compared with the control group 19 . In the present study, supplementation with EPA-PC and EPA-PE did not reverse the decrease of In Table 6 Table 5 The lipid profile of PC in cerebral cortex after dietary intake of EPA enriched phospholipids. Compared with glycerophospholipid, ether-linked phospholipid has more potent biological activity of brain func-tion, but the possible underlying mechanism has not been elucidated. Among alkyl ether analog of phosphatidylethanolamine ePE molecular species, 16:0e/22:4 and 18:0e/20:4 were the main fatty acids, as shown in Table 9, and there was no obvious change. What s more, the main fatty acids, such as 16:0e/16:0, 16:0e/16:1, 16:0e/18:1, 16:0e/22:1, and 18:0e/22:6, in alkyl ether analog of phosphatidylcholine ePC molecular species had no significant change as shown in Table 10.
Compared with the high-fat group, EPA-PC and EPA-PE diet reduced the relative content of DHA-containing pPC Table 11 . The potential mechanisms have not been illustrated yet and require further study.

PUFA content in glycerophospholipids
The main long chain fatty acids, including DHA, AA, docosapentaenoic acid DPA and EPA, in glycerophospholipids from cerebral cortex are important components for brain growth and function. DHA is mainly deposited in brain and retina, which is an important component of cell membrane 23 . DHA accounts for more than 90 of n-3 PUFAs and 10 -20 of total lipids in the brain 24 . Results showed that DHA exhibited the high levels in the molecular species of glycerophospholipid except for ePE, while there was no obvious change for DHA in PC, PE, pPE and PS in Fig. 2.
As an important source of n-3 PUFA, EPA has a lot of nutritional functions, such as reducing blood lipids, antiatherosclerosis, anti-thrombosis and supplement essential nutrients for the brain. In cerebral cortex of AD model mice, most of EPA was combined with PE and pPE Figs. 2B and C , which might be due to the responsibility of PE and pPE for the membrane fluidity. Although no significant changes were found in EPA-PC and EPA-PE groups compared with the HF group, notably, there was significant difference between EPA-PC and EPA-PE groups. Valentini et al. also reported that fish oil supplementation did not change the DHA and EPA levels in the brain 25 .
AA exists in the phospholipids of cell membrane in the form of esterification, and can be released from phospholipids through a variety of enzyme pathways. AA and its metabolites have strong physiological activity in the nervous and immune system. The levels of AA in PC, PE and ePE were significantly higher than those in other glyc-      erophospholipids including pPE and PS Fig. 2 . The intake of EPA-PE could significantly increase the levels of AAcontaining PE and PS in cerebral cortex, while EPA-PC supplementation could only enhance the level of DHA-containing PS. In addition, the administration with EPA-PC and EPA-PE could significantly reduce the levels of AAcontaining ePE in cerebral cortex.
DPA is the most abundant n-3 long-chain polyunsaturated fatty acid in the brain after DHA, which has special benefits for the neuro-protection 26 . DPA was mainly concentrated in PE, pPE and PS, as shown in Fig. 2. EPA-PE intake could significantly reduce the level of DPA-containing pPE and PS caused by high fat diet in cerebral cortex, while EPA-PC only reduced DPA-containing PS. It has been reported that DHA-PS supplementation could remarkably decrease the level of DPA-containing PS compared with the HF group 20 .
In the present study, no significant changes were observed in the major brain glycerophospholipids PC, PE, pPE, PS and ePE of SAMP8 mice after administration of EPA-enriched phospholipids for 8 weeks. Despite the relatively low abundance of LPE, HF diet reduced LPE content in the cerebral cortex compared with low-fat fed rats 20 . In the present study, dietary EPA-PC and EPA-PE significantly increased the LPE content in the cerebral cortex of highfat fed mice. The administration with EPA-PC and EPA-PE could not significantly change the levels of DHA and EPA in cerebral cortex, which might be attributed to the low dose of EPA-enriched phospholipids 1 , w/w in diet Tables 2 and 3 . In our previous study, 2 w/w EPA enriched phospholipids intervention significantly reduced the pathological symptoms of male SAMP8 mice, and changed the whole brain DHA and EPA level 27 . Compared with the low-fat group, high-fat diet could significantly decrease the levels of DHA-containing PS/pPE as well as AA-containing PE and increase the level of DPA-containing PS. Our results demonstrated that dietary EPA-PE could significantly reduce the level of DPA-containing pPE and PS caused by high fat diet in cerebral cortex, while EPA-PC reduced DPA-containing PS. Previous studies have shown that high consumption of butter or saturated fatty acids could aggravate the severity of age-related senile amyloidosis, while dietary 20 w/w soybean oil rich in n-6 polyunsaturated fatty acids for 21 days enhanced learning ability 28,29 . Moreover, compared with 5 w/w lard in diet, SAMP8 mice fed with 5 w/w soybean oil had better memory, greater longevity and higher concentrations of C16:0 and C18:2n-6 in the brain 30 .
The digestion and absorption of triacylglycerol TG and phospholipid PL in small intestine are different. PL is mainly hydrolyzed by pancreatic Phospholipase A 2 , which releases fatty acids from sn-2 position to form 1-lyso-PC 31 . After the released fatty acids and lyso-phospholipids enter enterocyte, a part of lyso-phospholipids is further hydro-lyzed into fatty acids and glycerophospholipids by an enterocyte-derived lysophospholipase. Moreover, the metabolism of lyso-phospholipids includes the re-esterification to TG 2-monoacylglycerol pathway as well as PL a-glycerophosphate pathway and the generation of chylomicrons for further transport 32,33 . The digestion of choline phospholipids is of great significance for choline homeostasis, lipid signaling, postprandial lipid and energy metabolism, and interaction with intestinal bacteria 34 . A previous study found that dietary PC was mainly hydrolyzed to 1-lyso-PC, which was effectively absorbed and reacylated to PC or degraded to glycerophoshocholine GPC , glycerophosphate and free choline 35 . Although dietary EPA-PC and EPA-PE can change the composition of fatty acids in the blood 36 , the brain is a closed system isolated by the blood-brain barrier, and supplementation of exogenous lipids may not cause significant changes in the composition of brain lipids.

Conclusion
Supplementation with EPA-PC and EPA-PE could improve brain function, and lipid homeostasis was closely related to human health. In the present study, the lipid profile in cerebral cortex was measured by lipidomics after treatment with EPA-PC and EPA-PE for 8 weeks in highfat diet induced SAMP8 mice with cognitive deficiency. Dietary supplementation with EPA-PE and EPA-PC for 8 weeks significantly increased the amount of pPC and LPE in the cerebral cortex of SAMP8 mice fed with high fat diet. Meanwhile, administration with EPA-PE and EPA-PC could significantly decrease the level of DPA-containing PS and increase the levels of AA-containing PE as well as PS in cerebral cortex. These results might provide a potential new treatment strategy and dietary intervention direction for patients with cognitive impairment.