The brain possesses a unique fatty acid composition with high levels of palmitate and the polyunsaturated fatty acids (PUFA), arachidonic acid (ARA, 20:4n-6), and docosahexaenoic acid (DHA, 22:6n-3), but low levels of other n-3 PUFA, especially eicosapentaenoic acid (EPA, 20:5n-3) (Svennerholm 1964; Crawford and Sinclair 1971; Brenna and Diau 2007). In recent years, EPA has been investigated as a therapy for several neurological diseases and disorders (Burgess et al. 2000; Peet et al. 2001; Peet and Horrobin 2002; Fux et al. 2004; Puri et al. 2005; Keck et al. 2006; Ross et al. 2007; Sorgi et al. 2007; Amminger et al. 2010). In the periphery, EPA and its oxygenated derivatives (eicosanoids and E-series resolvins) regulate immune function (Peterson et al. 1998; Thies et al. 2001; Hasturk et al. 2007; Ishida et al. 2010), decrease platelet aggregation (Ahmed and Holub 1984; Nieuwenhuys and Hornstra 1998; Adan et al. 1999), and increase cerebral blood flow (Black et al. 1984a,b; Katayama et al. 1997; Katsumata et al. 1999). However, the function of EPA in the brain is not clear and no study has reported the kinetics of EPA in the brain.
As endogenous synthesis of PUFA is low within the brain compared with uptake from the plasma unesterified fatty acid pool (Demar et al. 2005; DeMar et al. 2006), this suggests that the brain must maintain its PUFA concentrations via the uptake from plasma which may be supplied from the diet and/or the liver. Using an in situ cerebral perfusion competition assay, Ouellet et al. (2009) reported that 14C-EPA and 14C-DHA uptake were not saturable at supraphysiological concentrations and that 14C-EPA and 14C-DHA entered the brain at similar rates. Although these findings are consistent with the model proposed by Hamilton et al. (2001), it is unclear why brain phospholipids are enriched in DHA and low in EPA (Philbrick et al. 1987; Chen et al. 2008b; DeMar et al. 2008). As a follow-up, we found that EPA was 2.5-fold more readily β-oxidized than DHA, 40 s upon perfusion in situ (Chen et al. 2009). Although EPA β-oxidation was higher than DHA, this difference does not fully explain an up to 300-fold difference in their brain concentrations (Philbrick et al. 1987; Chen et al. 2008b; DeMar et al. 2008). This discrepancy may result from down-regulated β-oxidation in situ or the brain may be rapidly catabolizing de-esterified EPA from phospholipids.
With regards to saturated fatty acids, upon entry into the brain, palmitate is more readily β-oxidized as compared with ARA and DHA (Dhopeshwarkar and Mead 1969; Dhopeshwarkar et al. 1973; Sun and Horrocks 1973; Kawamura and Kishimoto 1981; Miller et al. 1987; Gnaedinger et al. 1988). Furthermore, de novo synthesis of palmitate is active in the brain (Carey 1975; Miller et al. 1987; Marbois et al. 1992; Lee et al. 1994a,b; Edmond et al. 1998); thus Jsyn may be a significant contributor of palmitate in brain phospholipids. In developing rats, Jsyn is 1992 nmol/g/day (Lee et al. 1994a); whereas in adult rats, Jin is 724–822 nmol/g/day (Grange et al. 1995; Chang et al. 1996; Contreras et al. 1999). Similar to ARA and DHA, if the plasma unesterified palmitate pool is a major plasma contributor to brain phospholipids, then the sum of Jsyn and Jin should approximate Jout. If Jout exceeds the sum of Jsyn and Jin, then it would suggest that other plasma fatty acid pools contribute to palmitate uptake into brain phospholipids. However, if the sum of Jsyn and Jin exceeds Jout, then it would imply that the known kinetic parameters may be overestimated or that the current model requires modification. In this study, we administered 14C-palmitate or 14C-EPA intracerebroventricularly to rats in order to calculate Jout for these fatty acids.
Materials and methods
- Top of page
- Materials and methods
- Supporting Information
All procedures were performed in accordance with the policies set out by the Canadian Council on Animal Care and were approved by the Animal Ethics Committee at the University of Toronto. Male Sprague Dawley rats were purchased from Charles Rivers (Saint-Constant, QC, Canada) at 12 weeks of age and kept at an animal facility with a 12 h light–dark cycle and a constant temperature of 22°C for 3 weeks. The rats received ad libitum access to standard chow (Teklad 2018; Harlan, Madison, WI, USA) and water. The linoleate (18:2n-6) and α-linolenate (18:3n-3) composition of the diet [measured by gas chromatography-flame ionization detection (GC-FID)] was 53% and 6%, respectively, and longer chain PUFA (20:3n-3, 20:4n-6, EPA, 22:4n-6, 22:5n-6, 22:5n-3 and DHA) were < 0.5%. At 15 weeks of age, seven rats were killed by head-focused, high-energy microwave irradiation (13.5 kW for 1.75 s; Cober Electronics Inc., Norwalk, CT, USA) to calculate baseline fatty acid concentrations and 45 rats were randomized to receive either 14C-palmitate ([1-14C]-palmitate, specific activity: 53 mCi/mmol; Moravek Biochemical Inc., Brea, CA, USA) or 14C-EPA ([1-14C]-EPA, specific activity: 54 mCi/mmol; Moravek Biochemical Inc.) intracerebroventricularly. The purity of the radiotracers was confirmed to be > 99% by HPLC and liquid scintillation counting (LSC).
Intracerebroventricular tracer infusion
Rats were anesthetised with isofluorane inhalation (3% induction, 1–2% maintenance) and then placed in a stereotaxic frame (Stoelting, Wood Dale, IL, USA). Before the incision was made, 50 μL of 0.5% sensorcaine was injected subcutaneously at the incision site. The skull was exposed and a small hole was drilled (+1.5 mm lateral/medial and −1 mm anterior/posterior from bregma). A 5 μL injection containing 10 μCi of 14C-palmitate (Gatti et al. 1986; DeMar et al. 2004; Green et al. 2010) or 14C-EPA dissolved in 5 mm HEPES buffer (pH 7.4) containing 50 mg/mL fatty acid-free bovine serum albumin was infused at a constant rate of 0.17 μL/min into the right lateral ventricle (−4 mm dorsal/ventral) of the brain using a 33-gauge beveled injection needle (World Precision Instruments, Sarasota, FL, USA). Thus, 185 nmol of palmitate and 189 nmol of EPA were injected over 29.4 min. Equivalent doses of radiolabeled palmitate and EPA were administered because the uptake rate of these fatty acids into the brain appears to be similar (Chen et al. 2008a; Ouellet et al. 2009). During the surgery, rats were placed on a heating pad and given a subcutaneous injection of 1 mL 0.9% NaCl. Five minutes after the infusion, the needle was removed at a rate of 1 mm/min. The skull was mended with cranioplastic cement and the incision was stitched with self-dissolving sutures. After surgery, rats recovered under a heating lamp for 20–30 min before returning to their cages. At 4, 16, 32, 64, and 128 days post-infusion, rats were killed by head-focused, high-energy microwave irradiation (13.5 kW for 1.75 s). Brains were excised and stored at −80°C until further biochemical analyses.
Total lipids were extracted according to the method of Folch, Lees, and Sloane Stanley using chloroform : methanol : 0.88% KCl (2 : 1 : 0.75 by vol.) (Folch et al. 1957). Isolation of various lipid classes from the total lipid extract was achieved by TLC. All TLC plates were washed in chloroform : methanol (2 : 1 by vol.) and activated by heating at 100°C for 1 h prior to use. Neutral lipids were separated with TLC G-plates (EMD Chemical, Gibbstown, NJ, USA) along with authentic standards (Avanti, Alabaster, AL, USA) in heptane : diethyl ether : glacial acetic acid (60 : 40 : 2 by vol.). TLC H-plates (Analtech, Newark, DE, USA) were used to separate phospholipid fractions along with authentic standards in chloroform : methanol : 2-propanol : 0.25% KCl : triethylamine (30 : 9 : 25 : 6 : 18 by vol.). Bands corresponding to authentic standards for total phospholipids, cholesterol, unesterified fatty acids, triglycerides, cholesteryl esters, ceramide phosphocholine, choline glycerophospholipids (ChoGpl), phosphatidylserine (PtdSer), phosphatidylinositol (PtdIns), and ethanolamine glycerophospholipids (EtnGpl) were visualized under UV light after spraying with 0.1% 8-anilino-1-naphthalene sulfonic acid.
Bands corresponding to total phospholipids and phospholipid fractions were collected into test tubes with a known amount of heptadecanoic acid (17:0), then converted to fatty acid methyl esters (FAME; ester-linked fatty acids) and fatty aldehyde dimethyl acetals (FADMA; vinyl ether-linked fatty acids) with 14% boron trifluoride-methanol at 100°C for 1 h. FAME from rats at 15 weeks of age (baseline) were quantified by GC-FID as described. FAME and FADMA from 14C-palmitate- and 14C-EPA-infused brains were separated by HPLC, counted by LSC and identified by gas chromatography-mass spectrometry (GC-MS) and GC-FID as described.
Cholesterol bands from neutral lipid TLC were collected and saponified in 1 m methanolic NaOH at 90°C for 1 h. Subsequently, the addition of saline and hexane separated the non-saponifiable cholesterol from saponifiable materials. After centrifugation at 275 g for 4 min, the upper hexane phase containing cholesterol was transferred to a separate test tube. The hexane wash and centrifugation were repeated to enhance yield. Cholesterol extracts were reconstituted and transferred into scintillation vials with 5 mL of scintillation cocktail and counted by LSC.
High performance liquid chromatography
FAME and FADMA from total and individual phospholipid pools were separated by HPLC (Waters 2690, Boston, MA, USA) equipped with an in-line UV photodiode array detector (Waters 996) set at a wavelength of 242 nm. Initial conditions were set at a 1 mL/min gradient system consisting of (i) 100% H2O and (ii) 100% acetonitrile. The gradient commenced with 85% (ii) for 30 min, then increased to 100% (ii) over a 10-min period where it was maintained for 20 min before returning to 85% (ii) over a 5-min period (Aveldano et al. 1983). Two columns were selected for total phospholipid separation (Luna C18, 4.6 × 250 × 5 μm; Phenomenex, Torrance, CA, USA and Symmetry C18, 4.6 × 250 × 5 μm; Waters, Mississauga, ON, Canada). Fractions from the Luna column were collected every minute for 55 min. The Luna column provided general separation of various saturates, monounsaturates, and polyunsaturates. Fractions from the Symmetry column were collected every 10 s from 26 to 46 min of the 55-min run. The Symmetry column separated peaks that co-elute with the Luna column (palmitate/oleate and palmitoleate/n-3 docosapentaenoate). Radioactivity was quantified by LSC and peak identity was confirmed by GC-FID and GC-MS.
The Symmetry column resolved the co-elusion of palmitoleate and n-3 docosapentaenoate but was not capable of separating palmitate and oleate baseline to baseline. Thus after palmitate/oleate fractions from the Symmetry column were collected, one-third of each fraction was removed for fatty acid quantification by GC-FID; while the other two-thirds were counted by LSC. Multiple linear regression analysis was performed to determine the relative radioactivity of palmitate and oleate.
Liquid scintillation counting
Total phospholipid and phospholipid fraction bands as well as HPLC fractions from radiotracer-infused brains were added to scintillation vials with 5 mL of scintillation cocktail (GE Healthcare Life Sciences, Baie d’Urfe, QC, Canada). Radioactivity was quantified by a Packard TRI-CARB2900TR liquid scintillation analyzer (Packard, Meriden, CT, USA) with a detector efficiency of 48.8%. Radioactivity was expressed in disintegrations per minute; then converted to nCi/brain.
Gas chromatography-flame ionization detection
Fatty acid methyl esters were analyzed using a Varian-430 gas chromatograph (Varian, Lake Forest, CA, USA) equipped with a Varian FactorFour capillary column (VF-23ms; 30 m × 0.25 mm i.d. × 0.25 μm film thickness) and a FID. Samples were injected in splitless mode. The injector and detector ports were set at 250°C. FAME were eluted using a temperature program set initially at 50°C for 2 min, increasing at 20°C/min, and held at 170°C for 1 min, then at 3°C/min and held at 212°C for 5 min to complete the run at 28 min. The carrier gas was helium, set to a constant flow rate of 0.7 mL/min. Peaks were identified by retention times of authentic FAME standards (Nu-Chek Prep, Inc., Elysian, MN, USA). The concentration of each fatty acid from phospholipids and unesterified fatty acids at baseline was calculated by comparison with the internal standard (17:0) and expressed as nmol/g brain (Chen et al. 2009).
Gas chromatography-mass spectrometry
Radioactive fractions and unesterified fatty acids were analyzed with an Agilent 6890 series gas chromatograph (Agilent Technologies, Wilmington, DE, USA) equipped with a HP-5 ms capillary column (Agilent Technologies; 30 m × 0.25 mm i.d. × 0.25 μm film thickness) and an Agilent 5973 Network Mass Selective Detector (Agilent Technologies). The sample was injected in splitless mode. The injector port, ion source, and interface were 250, 230, and 280°C, respectively. FAME and FADMA were eluted using a temperature program initially set at 50°C for 1 min, then increased at 15°C/min to 240°C and held for 5 min, and then increased at 15°C/min to 280°C for 8 min to complete the run in 29.33 min. The carrier gas was helium, set to a 1 mL/min constant flow. A mass range from 50 to 700 amu was scanned using an electron ionization energy of 70 eV. Peaks were identified on the basis of selected fragmented ions. Peak identification was also confirmed by their retention time via GC-FID.
Radioactivity of total phospholipids and phospholipid fractions were adjusted by the percentage of radiolabeled palmitate and EPA determined via HPLC and LSC. The 14C-palmitate and 14C-EPA radioactivity was log-transformed and plotted against day post-intracerebroventricular infusion, then the data were fitted by linear regression to provide slope (per day) (GraphPad Prism version 4.0, La Jolla, CA, USA).
Loss half-lives of 14C-palmitate and 14C-EPA in brain phospholipids were calculated from the slopes of the regression lines from total phospholipids and phospholipid fractions by the following equation (Stinson et al. 1991):
Then, the half-lives (t1/2) of palmitate and EPA were used to calculate their rate of loss (Jout, nmol/g brain/day) from brain phospholipids, by the following equation (Rapoport et al. 2001):
where CFA is the baseline fatty acid concentration of palmitate or EPA in a phospholipid pool (nmol/g brain).
The fractional loss was then calculated by the following equation (Green et al. 2010):
- Top of page
- Materials and methods
- Supporting Information
Similar to previous reports (Philbrick et al. 1987; Chen et al. 2008b; DeMar et al. 2008), palmitate was the most abundant fatty acid in brain total phospholipids at 21358 ± 609 nmol/g brain or 27% of total fatty acids. Upon adjustment for pool size, palmitate was predominately esterified to ChoGpl (46% of total ChoGpl fatty acid) as compared with EtnGpl (9.1%), PtdIns (14%), and PtdSer (2.4%). As expected, the concentration of EPA was relatively low in brain phospholipids at 116 ± 12 nmol/g brain or 0.15% of total fatty acids, which is comparable with the reported human brain EPA concentration of 122 ± 102 nmol/g brain or 0.13% of total fatty acids (Igarashi et al. 2010). Although EPA appeared to be selectively esterified to ChoGpl (20 ± 1.0 nmol/g brain) and PtdSer (17 ± 0.9 nmol/g brain), when the concentration of EPA was adjusted for the total fatty acid content of each phospholipid fraction, 0.45% of total PtdIns fatty acid was EPA; whereas 0.06%, 0.02%, and 0.17% of ChoGpl, EtnGpl, and PtdSer fatty acids were EPA, respectively. In addition, when the esterification pattern of radiotracers was analyzed, similar to the fatty acid percent composition, 14C-palmitate and 14C-EPA were predominantly esterified to ChoGpl (72%) and PtdIns (47%), respectively. Thus, consistent with our previous in situ investigation, EPA appears to be preferably esterified to PtdIns (Chen et al. 2009). This is of interest as PtdIns modulates several signaling cascades (Berridge et al. 1989) and is a candidate target of therapies for bipolar disorder (Jope et al. 1996; Silverstone et al. 2002; Ding and Greenberg 2003) where EPA may be efficacious (Frangou et al. 2006, 2007).
As fatty acids enter the brain, they can be esterified or metabolized. At day 4 post-infusion, 82 ± 26 nCi of 14C-palmitate and 2 ± 0.9 nCi of 14C-EPA were esterified to total phospholipids, the major brain fatty acid pool. Therefore, 99.2% and 99.9% of 14C-palmitate and 14C-EPA, respectively, disappeared upon entry into the brain. Albeit the concentration of infused EPA was higher than the unesterified fatty acid pool initially, after 4 days the amount of esterified radiolabeled palmitate and EPA was 1.52 and 0.039 nmol/brain or 0.007% and 0.03% of the phospholipid pool, respectively. To test if fatty acids were transferred from brain to peripheral tissues, the liver was analyzed for radioactivity. Similar to ARA and 20:1n-9 (Golovko and Murphy 2006), for both 14C-palmitate- and 14C-EPA-infused brains, 0.02% of radioactivity was detected in the liver at day 4 post-infusion suggesting that a negligible amount of the radiotracers left the brain intact. This implies that the disappearance of fatty acids from the brain occurred via metabolism rather than diffusion from brain to periphery.
Unlike the intracerebroventricular administration of radiolabeled DHA (DeMar et al. 2004) or ARA (Green et al. 2010) where the majority of radioactivity detected in brain phospholipids was intact DHA and ARA, respectively, infused 14C-EPA was metabolized into radiolabeled n-3 docosapentaenoate and DHA in brain phospholipids suggesting that the brain can synthesize longer chain PUFA using EPA as a precursor (Kaduce et al. 2008). However, it is important to note that the radioactivity in n-3 docosapentaenoate and DHA was 0.048% and 0.064% of the initial dose, respectively. Furthermore, the appearance of radiolabeled saturates and monounsaturates in 14C-EPA-infused brains suggests that EPA was also β-oxidized into acetyl-CoA which can enter various synthetic pathways, including fatty acid synthesis (Cunnane et al. 2003). In regards to 14C-palmitate-infused brains, radiolabeled stearate, palmitoleate, oleate, and 20:1n-9 could originate from desaturation and elongation of infused 14C-palmitate or the β-oxidation product of 14C-palmitate, radiolabeled acetyl-CoA. At day 4 post-infusion, cholesterol from both 14C-palmitate- (4 ± 1 nCi/brain) and 14C-EPA-infused (3 ± 1 nCi/brain) brains was radiolabeled; thus β-oxidation of palmitate and EPA were confirmed (Cunnane et al. 1994). In addition to metabolism via mitochondrial β-oxidation and desaturation/elongation, EPA may have been converted to oxygenated derivatives such as eicosanoids and E-series resolvins and future studies examining this are warranted.
Along with various radiolabeled saturates, monounsaturates and PUFA, radiolabeled fatty aldehydes, such as palmitaldehyde, olealdehyde, and octadecanal were identified. The appearance of radiolabeled fatty aldehyde suggested that fatty acids were incorporated into plasmalogens, which are specialized pools of ChoGpl and EtnGpl with vinyl ether-linked fatty acids rather than ester-linkages. Therefore, as expected, ChoGpl and EtnGpl fractions, but not PtdSer and PtdIns, had radiolabeled fatty aldehydes when individual phospholipid fractions were examined (Sun and Horrocks 1968; Diagne et al. 1984). In accordance with previous reports (Miller et al. 1987; Cunnane et al. 2006), the appearance of de novo synthesized radiolabeled fatty acids, fatty aldehydes, and cholesterol suggests that the brain conserved carbons from fatty acid β-oxidation.
In addition to the initial rapid disappearance of EPA in the brain, esterified 14C-EPA in brain total phospholipids was also rapidly lost via de-esterification. The EPA half-life of 5 days was 6–9 times faster than palmitate (half-life = 32 days), ARA (half-life = 44 days; Green et al. 2010) and DHA (half-life = 33 days; DeMar et al. 2004). Since de novo synthesis (Jsyn) of PUFA in the brain is negligible, the Jout of EPA (16 nmol/g/day) should approximate the rate of incorporation (Jin) of EPA from the plasma unesterified pool into brain phospholipids. This prediction could be tested with intravenous infusion of radiolabeled EPA.
The reported Jin of palmitate (724–822 nmol/g/day) is 1.5- to 1.8-fold higher than our calculated Jout for palmitate (469 nmol/g/day). However, the Jin for palmitate may be overestimated as it was assumed that total radioactivity detected in brain phospholipids was intact 14C-palmitate (Grange et al. 1995; Chang et al. 1996; Contreras et al. 1999). Since some fatty acids are rapidly metabolized by the brain as evident by linoleate (DeMar et al. 2006) and α-linolenate (Demar et al. 2005), where 5 min after intravenous infusion, 71% and 63% of radioactivity in brain phospholipids were fatty acid metabolites, respectively, it is possible that palmitate is also rapidly metabolized. Future studies correcting Jin for intact palmitate are warranted. Furthermore, because of recycling of 14C into de novo fatty acids, the half-life of infused 14C-palmitate quantified in this study may be underestimated. Future studies identifying the position of radiolabeled carbon could improve the estimation of Jout for palmitate (Corso and Brenna 1997; Huang et al. 2000). Once these kinetic parameters are established, the difference between Jin and Jout could be used to estimate the Jsyn of palmitate. Although, Jsyn of palmitate has been calculated to be 1992 nmol/g/day, this value was obtained from developing rats which have an increased requirement for palmitate because of myelination during brain growth (Dhopeshwarkar et al. 1969; Marbois et al. 1992); whereas mature rats have a lower palmitate requirement and down-regulated fatty acid synthetase activity in the brain (Cantrill and Carey 1975).
In conclusion, the low concentration of EPA in brain phospholipids appears to be the results of rapid metabolism, in part via β-oxidation, upon entry into the brain and the rapid de-esterification (t1/2 = 5 days) and fractional loss (14% per day) of phospholipid-esterified EPA. In contrast to EPA, palmitate is stably incorporated into brain phospholipids with a half-life (t1/2 = 32 days) and fractional loss (2% per day) comparable to ARA and DHA.
- Top of page
- Materials and methods
- Supporting Information
Figure S1. Identification and estimation of baseline unesterified EPA concentration in the brain
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.