These authors contributed equally to this work.
Docosahexaenoic acid synthesis from n-3 fatty acid precursors in rat hippocampal neurons
Article first published online: 4 FEB 2008
© 2008 The Authors. Journal Compilation © 2008 International Society for Neurochemistry
Journal of Neurochemistry
Volume 105, Issue 4, pages 1525–1535, May 2008
How to Cite
Kaduce, T. L., Chen, Y., Hell, J. W. and Spector, A. A. (2008), Docosahexaenoic acid synthesis from n-3 fatty acid precursors in rat hippocampal neurons. Journal of Neurochemistry, 105: 1525–1535. doi: 10.1111/j.1471-4159.2008.05274.x
Present Address: M.I.N.D. Institute, University of California at Davis, Sacramento, CA 95871, USA
- Issue published online: 4 FEB 2008
- Article first published online: 4 FEB 2008
- Received September 20, 2007; revised manuscript received January 9, 2008; accepted January 18, 2008.
- α-linolenic acid;
- arachidonic acid;
- eicosapentaenoic acid;
- linoleic acid;
Docosahexaenoic acid (DHA), the most abundant n-3 polyunsaturated fatty acid in the brain, has important functions in the hippocampus. To better understand essential fatty acid homeostasis in this region of the brain, we investigated the contributions of n-3 fatty acid precursors in supplying hippocampal neurons with DHA. Primary cultures of rat hippocampal neurons incorporated radiolabeled 18-, 20-, 22-, and 24-carbon n-3 fatty acid and converted some of the uptake to DHA, but the amounts produced from either [1-14C]α-linolenic or [1-14C]eicosapentaenoic acid were considerably less than the amounts incorporated when the cultures were incubated with [1-14C]22:6n-3. Most of the [1-14C]22:6n-3 uptake was incorporated into phospholipids, primarily ethanolamine phosphoglycerides. Additional studies demonstrated that the neurons converted [1-14C]linoleic acid to arachidonic acid, the main n-6 fatty acid in the brain. These findings differ from previous results indicating that cerebral and cerebellar neurons cannot convert polyunsaturated fatty acid precursors to DHA or arachidonic acid. Fatty acid compositional analysis demonstrated that the hippocampal neurons contained only 1.1–2.5 mol% DHA under the usual low-DHA culture conditions. The relatively low-DHA content suggests that some responses obtained with these cultures may not be representative of neuronal function in the brain.
- ALA or 18:3n-3
cytosine arabinoside C
- DHA or 22:6n-3
- EPA or 20:5n-3
glial fibrillary acidic protein
Docosahexaenoic acid (DHA or 22:6n-3), the most abundant n-3 polyunsaturated fatty acid present in the brain, is required for normal development and function of the nervous system (Lauritzen et al. 2001; Salem et al. 2001; Yavin 2006). DHA is contained in membrane phospholipids, primarily in the ethanolamine and serine phosphoglycerides fractions (Spector 1999), and it has a highly flexible structure that affects matrix and elastic properties of the lipid bilayer (Eldho et al. 2003; Gawrisch et al. 2003; Bruno et al. 2007). DHA promotes neuronal differentiation (Kawakita et al. 2006; Kan et al. 2007), prevents apoptosis (Kim et al. 2000; Florent et al. 2006), and activates neuronal gene expression by functioning as a ligand for the retinoid X receptor in the brain (Mata de Urquiza et al. 2000; Goldstein et al. 2003). In addition, DHA activates the Akt signaling pathway, a process that enhances neuronal survival (Akbar et al. 2005), and phospholipids containing DHA facilitate G protein-coupled signaling (Litman et al. 2001). Compounds synthesized from DHA also can affect neuronal function in various ways. For example, neuroprotectin D1, a 10,17S-docosatriene derivative of DHA formed by the lipoxygenase pathway, protects neural cells (Serhan et al. 2006; Bazan 2007), whereas neuroprostanes, which are isoprostane-like compounds produced from DHA during exposure to oxidative stress, damage the cells by altering the physical properties of the neural membranes (Roberts et al. 1998; Fam et al. 2002).
Because DHA is an essential fatty acid for mammalian cells, it must be supplied either directly to the cells or in the form of another n-3 fatty acid like α-linolenic acid (ALA or 18:3n-3) or eicosapentaenoic acid (EPA or 20:5n-3) that can be converted to DHA. The relative importance of these two mechanisms has been studied in some detail in astrocytes (Moore et al. 1991; Spector and Moore 1993; Williard et al. 2001a). However, little detailed information is currently available regarding neurons, and the purpose of the present study was to investigate this process in primary cultures of neurons isolated from the brain. Rat hippocampal neurons were selected as a model system because DHA has an important role in the hippocampus, a region of the brain that is critical for learning. Data from animal models indicate that a decrease in DHA content, which normally comprises 13% of the fatty acids in the hippocampus (Diau et al. 2005), significantly reduces spatial learning in rats as determined by the Morris water maze test (Lim et al. 2005a,b). This reduction is associated with a decrease in the cell body size of the CA1 pyramidal neurons (Ahmad et al. 2002). DHA also promotes neurite outgrowth in hippocampal neurons (Calderon and Kim 2004). Another reason for selecting hippocampal neurons as the model system is that these cultures are being widely used to investigate many important functional processes, including glutamatergic synapse formation (Jelks et al. 2007), endocannabinoid signaling (Ohno-Shosaku et al. 2007), β-amyloid-induced changes in synaptic vesicle recycling (Kelly and Ferreira 2007), the role of proinflammatory cytokines on synaptic network activity (Vereyken et al. 2007), and the role of dendritic contact in the assembly of the functional pre-synaptic apparatus (Krueger et al. 2003). Moreover, the low-density serum-free defined medium allows healthy neuronal maturation with minimal glia cell growth, enabling us to study fatty acid metabolism in highly enriched neuron cultures.
Most of the available in vivo data indicate that the brain primarily utilizes DHA obtained directly from the diet or produced in the liver from dietary n-3 fatty acid precursors (Scott and Bazan 1989; Su et al. 2000; Rapoport et al. 2001b; DeMar et al. 2005; Lefkowitz et al. 2005). This mechanism is supported by previous results demonstrating that neurons cultured from the forebrain gray matter of 2- to 7-day-old rat pups took up DHA but did not synthesize DHA from ALA or EPA (Moore et al. 1991; Spector and Moore 1993). Likewise, cultures composed of 90% glutaminergic granular neurons and 5–7% GABAergic neurons obtained from the cerebellum took up DHA but did not convert ALA or EPA to DHA (Moore et al. 1991). The cerebral and cerebellar neuron cultures also did not convert linoleic acid (18:2n-6) to arachidonic acid (20:4n-6), the main n-6 fatty acid in the brain, providing additional evidence that neurons cannot convert polyunsaturated fatty acid precursors to the highly unsaturated derivatives required by the brain (Moore et al. 1991). As opposed to these previous results, we now find that primary cultures of rat hippocampal neurons express the complete polyunsaturated fatty acid metabolic pathway (Sprecher et al. 1995), and they are able to convert n-3 fatty acid precursors to DHA and linoleic acid to arachidonic acid.
Primary culture of rat hippocampal neurons
Low-density cultures of dissociated rat hippocampal neurons were prepared using previously described methods (Lim et al. 2003). In brief, hippocampi were dissected from 18-day embryonic rats (Harlan Sprague–Dawley, Indianapolis, IN, USA). The pooled tissue was incubated in HEPES-buffered saline solution (Ca2+/Mg2+-free) containing 0.03% trypsin at 37°C for 15 min. After the tissue was washed, it was dissociated with a fire-polished Pasteur pipette and plated at a density of 3–6 × 103/cm on coverslips (Warner Instruments, Hamden, CT, USA) coated with 0.1% (w/v) poly-l-lysine (Peptides International, Louisville, KY, USA). This low-density culture was plated in Neurobasal media containing B-27 supplement (prepared in our laboratory according to a published formulation; Brewer et al. 1993), 0.5 mM glutamine, and 5% horse serum. This medium was replaced with serum-free Neurobasal medium supplemented with B-27 and 0.5 mM glutamine 4 h after plating. Cells were maintained at 37°C in a humidified environment of 95% air and 5% CO2. In some experiments, cultures maintained for 7 days were treated overnight with 2 μM cytosine arabinoside hydrochloride (Ara-C; Sigma, St Louis, MO, USA) to remove astrocytes, and the cultures were then continued in normal growth medium. Radiolabeled DHA formation was compared in cultures treated with Ara-C and corresponding untreated cultures.
Fluorescence microscopy, imaging, and antibodies
The purity of the hippocampal cultures was determined by immunofluorescence microscopy using a Olympus IX-70 inverted epifluorescence microscope (Leeds, Minneapolis MN, USA) equipped with an Olympus (Center Valley, PA, USA) 10× dry objective, a MAC2002 Ludl shutter (BD Biosciences Bioimaging, Rockville, MD, USA), and fluorescence filter sets (Chroma, Brattleboro, VT, USA) for fluorescein (490 nm band-pass excitation and 528 nm long-pass emission) and rhodamine (555 nm band-pass excitation and 617 nm long-pass emission). Images were acquired with an ORCA II CCD camera (Hamamatsu, Bridgewater, NJ, USA) equipped with a frame grabber EDT DV PCI card controlled by Esee software (Inovision, Chapel Hill, NC, USA). Coverslips on which hippocampal cells were grown were rinsed in Dulbecco’s phosphate-buffered saline (Invitrogen, Carlsbad, CA, USA) before fixation. The coverslips were transferred to 4%p-formaldehyde with 4% glucose in the phosphate buffer for 15 min. Following fixation, the cells were permeabilized with 0.1% Triton X-100 for 20 min and incubated in phosphate-buffered blocking solution (2% glycerol, 0.05 M NH4Cl, 5% fetal bovine serum, and 2% donkey serum) for 2 h. Primary antibodies were added, and the cells were incubated for either 1.5 h or overnight at 4°C. The cells were again washed and incubated for 1 h at 23°C with Alexa 488 and 568 conjugated secondary antibodies (Molecular Probes, Eugene, OR, USA) diluted 1 : 100. After a final wash, the cells were mounted in Prolong Antifade mounting media before viewing. Antibodies used included a mouse monoclonal anti-microtubule-associated protein (MAP) 2B (1 : 100; Transduction laboratories, Lexington, KY, USA) as a neuronal marker, and a rabbit polyclonal anti-glial fibrillary acidic protein (GFAP) (1 : 1000; Dako Cytomation, Glostrup, Denmark) to define astrocytes. For each individual experiment, 8–10 fields were randomly selected from both Ara-C treated and untreated coverslips. MAP2B- and GFAP-positive neurons were manually counted for quantification and statistics.
Incubation and lipid extraction
The growth medium was removed, and after washing the cells once with the phosphate-buffered saline solution, the cultures which contained an average of 200 μg cell protein were incubated in 1.5 mL serum-free Neurobasal medium containing B-27, 0.5 mM glutamine, supplemental fatty acid where indicated, and 0.1 μM bovine serum albumin (essentially fatty acid-free) for 24–72 h at 37°C in an humidified atmosphere containing 5% CO2. Radiolabeled fatty acids were purchased from American Radiolabeled Chemicals Inc. (St Louis, MO, USA); [1-14C]18:2n-6, [1-14C]18:3n-3, and [1-14C]docosapentaenoic acid ([1-14C]22:5n-3), 50–60 mCi/mmol; and [1-14C]20:5n-3 and [1-14C]22:6n-3, 40–60 mCi/mmol. The corresponding unlabeled fatty acids were obtained from Cayman Chemical (Ann Arbor, MI, USA). Dr Howard Sprecher, Ohio State University, kindly provided [3-14C]tetracosapentaenoic acid ([3-14C]24:5n-3), [3-14C]tetracosahexaenoic acid ([3-14C]24:6n-3), and the corresponding unlabeled fatty acids. The radiolabeled fatty acids were purified by HPLC, and the purity of unlabeled fatty acids was checked by GLC.
The incubations were terminated by removing the medium, washing the cells with ice-cold phosphate-buffered saline solution, and extracting the cell lipids with a 2 : 1 chloroform : methanol mixture. Following isolation of the chloroform phase and removing the solvent under a steam of N2, the lipid extract was dissolved in 0.2 mL chloroform and aliquots were taken for the various lipid analytical procedures. Fatty acid methyl esters were prepared from aliquots of the cell lipid extract by incubating the lipid dissolved in 0.5 mL acetonitrile with 0.25 mL 14% BF3 in methanol at 95°C for 45 min. After addition of 4 mL water, the methyl esters were extracted three times with 1 mL n-heptane. The n-heptane extracts were combined, the solvent removed under a stream of N2, and the methyl esters suspended in 50 μL of acetonitrile (Williard et al. 2001b).
The radiolabeled fatty acid methyl esters were separated by reverse phase HPLC using a gradient system consisting of water acidified to pH 3 with formic acid and acetonitrile at a flow rate of 0.7 mL/min (Williard et al. 2001a). The gradient started at 76% acetonitrile and increased to 100% acetonitrile at 50 min. After the column effluent was mixed with three volumes of Budget Solve scintillation solution (Research Products International Corp., Mt. Prospect, IL, USA), radioactivity was measured by passing the mixture through an in-line flow radioactivity detector (IN/US Inc., Tampa, FL, USA). Radiolabeled fatty acid methyl ester standards were included with each set of chromatograms.
Unlabeled fatty acid methyl esters were separated by GLC (Williard et al. 2001b). Margaric acid (17:0) was added as an internal standard prior to extraction. The separation was performed with a Hewlett-Packard 5890 chromatograph (Palo Alto, CA, USA) containing a 1.9 m × 2 mm glass column packed with GP-10% SP2330 on 100–120 mesh Chromosorb WAW (Supelco, Bellefonte, PA, USA). The carrier gas was N2 at a flow rate of 25 mL/min. Fatty acid methyl esters were detected by flame ionization at 250°C, identified by comparison with the retention times of fatty acid methyl ester standards, and quantified by peak area integration.
Cell lipids were separated by TLC using Whatman LK5D plates (Whatman Inc., Florham Park, NJ, USA) and a solvent mixture containing chloroform : methanol : methylamine : water (60 : 36 : 1.5 : 1). After development and drying, the plates were scanned for radioactivity with a Bioscan TLC plate gas proportional flow detector (Bioscan, Washington, DC, USA). Phospholipid standards (Avanti Polar Lipids, Birmingham, AL, USA) were added to each TLC plate. After scanning, the dried TLC plate was sprayed with 1 mM 8-anilino-1-naphthalene and the lipid standards were visualized under ultraviolet light. The radiolabeled lipids were identified by co-migration with the lipid standards (Vartak et al. 1998).
All data are presented as mean ± SD. Differences between mean values were analyzed by Student’s t-test, and a value of p < 0.05 were considered to be statistically significant.
Comparative incorporation of n-3 fatty acids
The initial studies compared the uptake of [1-14C]18:3n-3, [1-14C]20:5n-3, and [1-14C]22:6n-3 by the cultured hippocampal neurons. Figure 1 shows that the cells took up each of these fatty acids. When the medium containing 4 μM fatty acid (Fig. 1a), 2.6 times more 22:6n-3 was incorporated into the cell lipids than 18:3n-3, but there was no significant difference between the 20:5n-3 and 22:6n-3 uptake. More of each fatty acid was incorporated when the culture medium contained 10 μM fatty acid, but there was still a twofold difference in the amount of 22:6n-3 taken up when compared with 18:3n-3 (Fig. 1b).
Docosahexaenoic acid utilization
Table 1 contains results of TLC analysis of the lipids extracted from cells incubated for 66 h with either 4 or 10 μM [1-14C]22:6n-3. Between 90% and 95% of the radioactivity taken up was present in phospholipids, and the amount incorporated into each phospholipid fraction was 30–60% greater in the incubation with 10 μM [1-14C]22:6n-3. The ethanolamine phosphoglycerides contained from 50% to 59% of the incorporated radioactivity, choline phosphoglycerides contained 17–28%, and the combined serine and inositol phosphoglyceride fraction contained 15–18% (n = 6 separate cultures). In agreement with these results, recent studies with PC12 neurons also demonstrated that more DHA was incorporated into ethanolamine phosphoglycerides than any of the other phospholipids (Richardson and Wurtman 2007).
|Phospholipid fraction||Radiolabeled DHA content (nmol/200 μg cell protein ± SD)b|
|4 μM [1-14C]22:6n-3c||10 μM [1-14C]22:6n-3c|
|PtdSer + PtdIns||0.62 ± 0.035||0.82 ± 0.076d|
|PtdCho||0.73 ± 0.210||1.08 ± 0.028|
|PtdEtn||1.98 ± 0.189||3.16 ± 0.133d|
Additional studies were performed to determine whether the DHA incorporated by the hippocampal cells underwent structural modification. Aliquots of the cell lipid extract were methylated and the radiolabeled fatty acid methyl esters were analyzed by HPLC. The results, which are shown in Fig. 2, demonstrate that more than 99% of the incorporated radioactivity remained as unmodified 22:6n-3 in cells that were maintained in culture for either 11 days (Fig. 2a) or 18 days (Fig. 2b) prior to the 24 h incubation with [1-14C]22:6n-3.
Conversion of n-3 fatty acid precursors to DHA
Radiolabeled DHA was produced when the hippocampal cultures were incubated with n-3 fatty acid precursors. Figure 3 contains representative HPLC obtained from cultures incubated for 72 h with 5 μM [1-14C]18:3n-3 (Fig. 3a), [1-14C]20:5n-3 (Fig. 3b), [1-14C]22:5n-3 (Fig. 3c), [3-14C]24:5n-3 (Fig. 3d), and [3-14C]24:6n-3 (Fig. 3e). Quantitative results from these and other cultures that were tested are contained in Table 2. In the 72 h incubations with [1-14C]18:3n-3, 27% of the radioactivity in the cells was present in 22:6n-3, 26% in 22:5n-3, 32% in 20:5n-3, 9% in palmitic acid (16:0), and 0.4% in stearic acid (18:0). Only trace quantities of radiolabeled 18:3n-3 remained in the hydrolyzed cell lipid extract. In the single culture incubated for 72 h with [1-14C]20:5n-3, 50% of the radioactivity was present in 22:6n-3 and 26% in 22:5n-3, 8% remained as 20:5n-3, and 1.5% in 16:0. In the 72 h incubations with 5 μM [1-14C]22:5n-3, 36% of the radioactivity was present in 22:6n-3, 21% in 16:0, 5% in 18:0, and 33% remained as unmodified 22:5n-3. Results from 24 h incubations with [1-14C]18:3n-3 and [1-14C]20:5n-3 demonstrate that the conversion of these fatty acids to 22:6n-3 was time dependent. However, the amounts of radiolabeled 22:6n-3 produced from [1-14C]18:3n-3 and [1-14C]20:5n-3 were 75% and 55% less, respectively, than the amount present in the cell lipids after incubation for 24 h with [1-14C]22:6n-3. The standard deviations of the results obtained with these fatty acids are very large, presumably because of differences in the metabolic status of the cultures.
|Radiolabeled fatty acid substratea||Time of incubation (h)||nb||n-3 fatty acid products (nmol/200 μg cell protein ± SD)c|
|[1-14C]18:3n-3||24||5||0.383 ± 0.216||0.492 ± 0.043||0.195 ± 0.225|
|[1-14C]18:3n-3||72||9||0.549 ± 0.409||0.453 ± 0.088||0.468 ± 0.397|
|[1-14C]20:5n-3||24||4||0.522 ± 0.050||0.688 ± 0.126||0.399 ± 0.090|
|[1-14C]22:5n-3||72||8||ND||1.916 ± 1.245||1.936 ± 0.766|
|[1-14C]22:6n-3||24||6||ND||ND||0.857 ± 0.266|
|[1-14C]22:6n-3||72||3||ND||ND||4.450 ± 0.176|
In the 72 h incubations with [3-14C]24:5n-3, 53% of the radioactivity was present in 22:6n-3, 31% in 16:0, 5% in 18:0, and 12% remained as unmodified 24:5n-3. The only appreciable radiolabeled product detected in the 72 h incubation with [3-14C]24:6n-3 was 22:6n-3, accounting for 93% of the incorporated radioactivity. This value is only 15% less that the amount of radiolabeled 22:6n-3 present in the cell lipids after a 72 h incubation with [1-14C]22:6n-3.
Fatty acid composition
GLC analysis was performed to determine whether the radiolabeled fatty acid data reflected actual changes in the fatty acid composition of the hippocampal neurons.
Results from two separate experiments are shown in Table 3. Control cultures maintained in media without fatty acid supplementation contained only 1.1–2.5 mol% 22:6n-3 in the cell lipids. Trends toward higher 20:5n-3, 22:5n-3, and 22:6n-3 mol% in the incubation with EPA, and toward higher 22:6n-3 mol% in the incubation with DHA were noted in the first experiment, but the differences were not statistically significant. While none of the differences between the control cultures and those supplemented with ALA were statistically significant in the second experiment, there was a 3.8-fold increase in 22:6n-3 mol% and a 65% decrease in 22:5n-3 mol% in the cultures supplemented with DHA (p < 0.05). Therefore, the amounts of DHA produced in the studies with either [1-14C]18:3n-3 or [1-14C]20:5n-3 were not large enough to appreciably increase in 22:6n-3 content of the cell lipids. However, an increase in 22:6n-3 mol% occurred in the incubations with DHA, and this is consistent with the larger radiolabeled DHA incorporation observed in Table 2 when the cultures were incubated with [1-14C]22:6n-3 when compared with either [1-14C]18:3n-3 or [1-14C]20:5n-3.
|Fatty acid||Cell fatty acid composition (mol% ± SD)|
|Experiment 1a||Experiment 2b|
|Saturated||55.9 ± 9.0||56.2 ± 8.0||56.0 ± 9.1||44.8 ± 1.9||47.0 ± 1.2||45.9 ± 5.4|
|Monounsaturated||23.9 ± 6.8||21.9 ± 7.3||23.2 ± 7.7||29.1 ± 2.8||26.7 ± 1.0||31.5 ± 4.3|
|n-6 polyunsaturated||15.2 ± 2.5||12.8 ± 3.0||14.1 ± 1.5||21.1 ± 1.7||19.1 ± 0.9||16.9 ± 2.1|
|20:4n-6||6.7 ± 1.5||4.7 ± 2.8||6.3 ± 0.6||7.8 ± 2.1||8.5 ± 0.7||6.4 ± 0.9|
|n-3 polyunsaturated||4.8 ± 1.0||8.9 ± 4.3||6.0 ± 1.1||5.0 ± 1.2||7.2 ± 1.2||5.7 ± 1.2|
|18:3n-3||1.0 ± 0.6||0.6 ± 0.4||0.4 ± 0.2||1.2 ± 1.4||0.7 ± 0.2||0.5 ± 0.2|
|20:5n-3||0.3 ± 0.2||2.4 ± 3.2||0.3 ± 0.1||0.7 ± 0.4||1.4 ± 0.4||0.3 ± 0.2|
|22:5n-3||1.0 ± 0.2||2.4 ± 1.2||0.9 ± 0.2||3.1 ± 0.7||3.9 ± 0.5||1.1 ± 0.2c|
|22:6n-3||2.5 ± 1.2||3.5 ± 1.2||4.5 ± 1.2||1.1 ± 0.2||1.9 ± 1.6||4.2 ± 0.9c|
Conversion of linoleic to arachidonic acid
Previous studies indicated that primary cultures of rat cerebellar and cerebral neurons did not convert linoleic acid to arachidonic acid (Moore et al. 1991; Spector and Moore 1993). However, the finding that radiolabeled ALA was converted to EPA suggested that the hippocampal neurons also might have the ability to synthesize arachidonic acid from linoleic acid. To evaluate this possibility, HPLC analysis of the methylated cell lipid extract was performed following incubations with radiolabeled linoleic acid (Fig. 4). In cultures incubated for 24 h with 5 μM [1-14C]18:2n-6, 84% of the incorporated radioactivity remained as 18:2n-6, 13% was converted to 20:4n-6, and 3% was recovered in 16:0 (Fig. 4a). In 72 h incubations, only 28% of the incorporated radioactivity remained as 18:2n-6, 47% was converted to 20:4n-6, 0.7% to eicosatrienoic acid (20:3n-6), and 6.7% to 16:0 (Fig. 4b). An unidentified component with a retention times of 46.6 min contained 6.8% of the incorporated radioactivity.
Tests for glial contamination
Previous studies indicated that primary astrocyte cultures prepared from neonatal rat brain converted large quantities of ALA and EPA to DHA (Moore et al. 1991; Williard et al. 2001a). To determine whether the conversions observed in the hippocampal cultures might be because of astrocyte contamination, hippocampal cultures were double stained for neuronal and glial markers. Figure 5 shows typical results obtained with the MAP2B neuronal marker (red fluorescence) and GFAP glial marker (green fluorescence). In some of the fields examined, no cells stained positive for glial markers (Fig. 5a); in others, one cell was positive (Fig. 5b). The results obtained with representative double stained cultures indicate that less than 5% of all cells were positive for glial markers. To further evaluate the possibility that glial contamination contributed to the amount of DHA produced from n-3 fatty acid precursors, studies were performed in which one set of the hippocampal cultures was treated overnight with 2 μM Ara-C to inhibit the growth of astrocytes. Table 4, which summarizes quantitative data obtained from three separate experiments, demonstrates that the glial content was small and that the composition of the cultures was not affected by Ara-C treatment. The modest effect of Ara-C most likely is due to the already very low level of astrocyte contamination of these cultures, making it difficult to achieve any further decrease.
|Cultures||Cell typea||Percent glia|
|Cells/microscopic field (Mean ± SD)||Mean % ± SD|
|Control||14.77 ± 1.28||0.49 ± 0.08||3.13 ± 0.86|
|Ara-C treatedd||14.46 ± 0.73||0.44 ± 0.06||2.80 ± 0.34|
In additional experiments, the production of radiolabeled fatty acid products was compared in control cultures and those treated with Ara-C. Both sets of cultures were incubated with 2 μM [1-14C]20:5n-3 for 48 h. No appreciable difference in incorporation of radiolabeled 20:5n-3 was observed; 1.63 ± 0.165 nmol/200 μg cell protein in the controls when compared with 1.56 ± 0.027 in the Ara-C-treated cultures (mean ± SD, n = 3). Likewise, as shown in Fig. 6, HPLC analyses of the methylated cell lipid extracts demonstrated that there were no appreciable differences in the amount of [1-14C]20:5n-3 elongated and desaturated by these cultures. The control cultures converted 43% and 30% of the incorporated radioactivity to 22:5n-3 and 22:6n-3, respectively (Fig. 6a), while those treated with Ara-C converted 44% and 29% to 22:6n-3 and 22:5n-3 (Fig. 6b). These findings, together with the immunofluorescence results, indicate that the synthesis of DHA from n-3 fatty acid intermediates most likely is not because of glial contamination of the hippocampal cultures.
These results demonstrate that rat hippocampal neuron cultures can synthesize DHA from n-3 fatty acid precursors. DHA is the main n-3 fatty acid in the brain and is essential for optimum neural function. All of the n-3 fatty acids tested were converted to DHA, including ALA, EPA, and the 22- and 24-carbon intermediates, indicating that the hippocampal neurons express the complete pathway for n-3 fatty acid conversion to DHA (Sprecher et al. 1995). Similar amounts of radiolabeled 22:6n-3 were produced in 72 h from 20:5n-3, 22:5n-3, and 24:5n-3, but five times less DHA was produced from ALA. By contrast, almost twice as much radiolabeled 22:6n-3 was produced from 24:6 than from 20:5n-3, 22:5n-3, or 24:5n-3. These results suggest that the Δ6 desaturase, which acts on both the 18:3n-3 and 24:5n-3 desaturation steps in DHA synthesis (Sprecher et al. 1995), most likely is the rate-limiting enzyme of this pathway in hippocampal neurons.
These results differ from our previous findings indicating that cultured rat cerebral and cerebellar neurons cannot convert ALA and EPA to DHA (Moore et al. 1991; Spector and Moore 1993). The reason for this difference is uncertain, but the comparative results from cultures treated with Ara-C demonstrate that it is not because of glial contamination of the neuronal cultures. DHA has many important functions in the hippocampus (Ahmad et al. 2002; Calderon and Kim 2004; Lim et al. 2005a,b; Lukiw et al. 2005), and the ability to synthesize DHA from ALA and EPA may insure that at least some DHA is available for hippocampal function if the external supply is limited, or it may provide a small amount of DHA for intracellular processes that cannot utilize externally available DHA. Alternatively, the disparity between the present findings and our previous results might be because of differences in the culture conditions. In addition to differences in the plating media, the hippocampal neurons were maintained in serum-free Neurobasal medium supplemented with B-27 and 0.5 mM glutamine (Brewer et al. 1993; Lim et al. 2003), while in our previous studies, the cerebral and cerebellar neurons were maintained in Eagle’s minimum essential medium containing 33 mM glucose, 80 μM fluorodeoxyuridine, 2.5% chick embryo extract, and 10% fetal bovine serum (Moore et al. 1991).
Arachidonic acid is the most abundant n-6 essential fatty acid present in the brain and, like DHA, it is necessary for optimum neural function. Our previous studies indicated that cerebral and cerebellar neuron cultures did not synthesize arachidonic acid from linoleic acid and, therefore, required a preformed source of arachidonic acid (Moore et al. 1991; Spector and Moore 1993). As opposed to these results, the present findings demonstrate that hippocampal neurons readily convert linoleic acid to arachidonic acid. This is consistent with the fact that mammalian cells contain a single polyunsaturated fatty acid desaturation and elongation pathway that utilizes both n-3 and n-6 fatty acids (Sprecher et al. 1995).
Humans, including infants, are known to convert ALA to DHA (Salem et al. 1996; Brenna 2002Burdge 2006). Studies in animal models indicate that this conversion occurs primarily in the liver and that the DHA is stored there and in other tissues prior to transfer to the brain (Scott and Bazan 1989; Lefkowitz et al. 2005; Igarashi et al. 2007). However, the conversion of ALA to DHA also can take place in the developing brain (Dhopeshwarkar and Subramanian 1976; Green and Yavin 1993; Su et al. 2000). Like the cerebral and cerebellar neurons, cultured brain microvessel endothelial cells are unable to convert ALA or EPA to DHA (Moore et al. 1990). Therefore, astrocytes were thought to be the cells where this process occurred, but the present findings demonstrate that DHA synthesis from n-3 fatty acid precursors also can occur in cultured neurons. Because of differences in the radiolabeled fatty acid concentrations and incubation conditions in the present and previous studies, it is not possible to estimate the relative contributions of the astrocytes and neurons to this process.
While results in rats and guinea pigs show that ALA can supply DHA to the nervous system, these studies indicate that dietary DHA is necessary for optimal DHA accretion (Woods et al. 1996; Abedin et al. 1999). Likewise, more DHA accumulates in glial cell ethanolamine and serine phosphoglycerides when the cultures are supplemented with DHA (Raffick et al. 2005). Kinetic studies in rats indicate that the daily uptake of [1-14C]22:6n-3 by the adult rat brain is 370–1160 nmol/g, a quantity sufficient to supply the daily DHA requirement (Rapoport et al. 2001b), whereas only 1% of the [1-14C]18:3n-3 taken up by the rat brain was converted to DHA (DeMar et al. 2005; Igarashi et al. 2007). These in vivo studies were performed with adult rats, whereas the present data were obtained with neurons cultured from 18-day embryonic rats. Likewise, the astrocytes studied previously were cultured from 2- to 7-day-old rat pups (Moore et al. 1991; Williard et al. 2001a). A 12-fold decline in fatty acid Δ6-desaturase activity occurs in the brain during the 21-day postnatal period (Bourre et al. 1990). This decrease may account for the larger conversions of ALA to DHA in the cultured astrocytes and neurons than would be predicted from the in vivo studies in the adult rat brain (DeMar et al. 2005; Igarashi et al. 2007).
Notwithstanding these differences, the present results still are consistent with the conclusion that DHA obtained from the plasma is the main source of DHA for the brain (Rapoport et al. 2001b; DeMar et al. 2005, 2006). Rat brain contains 15 300 nmol/g of esterified DHA, and the daily turnover is between 2% and 8%, or 306 and 1224 nmol/g (Rapoport et al. 2001b). From the results in Table 2, the conversion of 5 μM [1-14C]18:3n-3 to DHA by the neurons is 146 nmol/g cells in 24 h. This suggests that the conversion of ALA to DHA in the neurons can account for only 12–48% of the daily DHA requirement of the adult rat brain. Similar calculations indicate that the hippocampal neurons incubated with 5 μM [1-14C]22:6n-3 incorporated 644 nmol/g cells in 24 h, a value within the range of daily DHA turnover in the rat brain (Rapoport et al. 2001b). Furthermore, this value is 4.4 times larger than the calculated conversion of [1-14C]18:3n-3 to DHA under these conditions. This difference probably accounts for the fact that as opposed to ALA, supplementation of the cultured neuron with DHA for 72 h significantly increased the 22:6n-3 mol% of the cell lipids.
Recycling of ALA carbons for fatty acid synthesis through β-oxidation has been observed in suckling rats and rhesus monkeys (Cunnane et al. 1999), and kinetic studies in rats indicate that almost all of the [1-14C]18:2n-6 or [1-14C]18:3n-3 taken up by the brain undergoes β-oxidation (DeMar et al. 2005, 2006). The hippocampal neuron cultures produced radiolabeled palmitic acid, the main product of fatty acid de novo synthesis, from [1-14C]18:2n-6, [1-14C]18:3n-3, [1-14C]22:5n-3, and [3-14C]24:5n-3. A small amount of radiolabeled stearic acid also was produced from [1-14C]22:5n-3 and [3-14C]24:5n-3. This is most likely because of fatty acid β-oxidation, followed by recycling of the resulting acetyl-CoA. The functional significance of this recycling process is unknown. It may be related to the finding that saturated fatty acids do not enter the brain (Noelle Marbois et al. 1992; Edmond et al. 1998; Edmond 2001), suggesting that one function of polyunsaturated fatty acids uptake may be to provide the acetyl-CoA needed by the brain for non-essential fatty acid synthesis and fatty acid chain elongation. However, this explanation seems unlikely because other studies show that the brain can take up saturated fatty acids from the plasma. For example, [9,10-3H]palmitic acid rapidly crosses the blood–brain barrier in rats and is incorporated into brain phospholipids (Grange et al. 1995; Contreras et al. 1999; Rapoport 2001a), and enough [1-11C]palmitic acid injected intravenously enters the brain for positron emission tomography imaging in monkeys (Arai et al. 1995).
Possibly, the most important result of the present work is that hippocampal neurons cultured in Neurobasal medium supplemented with B-27, which are being used extensively as a model system for studies of neuronal function, contain only 1.1–2.5 mol% DHA. By contrast, brain fatty acid compositional analysis indicates that the hippocampus ordinarily contains 13% DHA, and most other regions of the brain contain between 11% and 15% (Diau et al. 2005). The low-DHA content is likely because of the fact that the cultures were plated in horse serum, which contains <0.1% DHA (Stoll and Spector 1984), and then grown in serum-free Neurobasal medium. Although the B-27 supplement contains 3.6 μM ALA, the results in Table 3 show that this amount of ALA is not sufficient to raise the DHA content of the cell lipids to physiological levels. Thus, it is possible that some functional responses obtained with these widely used hippocampal neuron cultures may be suboptimal or even impaired because of their relatively low-DHA content.
These studies were supported by research Grants R01 HL072845 from the National Heart Lung and Blood Institute (AAS) and R01 NS046450 from the National Institute for Neurological Diseases and Stroke (JWH), National Institutes of Health. We thank Dr Howard Sprecher, Ohio State University, for providing the 24-carbon n-3 fatty acids and isotopes.
- 1999) The effects of dietary α-linolenic acid compared with docosahexaenoic acid on brain, liver and heart in the guinea pig. Lipids 34, 475–482. , , and (
- 2002) A decrease in cell size accompanies a loss of docosahexaenoate in the rat hippocampus. Nutr. Neurosci. 5, 103–113. , , , and (
- 2005) Docosahexaenoic acid: a positive modulator of Akt signaling in neuronal survival. Proc. Natl Acad. Sci. USA 102, 10858–10863. , , and (
- 1995) Incorporation of [1-carbon-11]palmitate in monkey brain using PET. J. Nuc. Med. 36, 2261–2267. , , , , , , , , and (
- 2007) Omega-3 fatty acids, pro-inflammatory signaling and neuroprotection. Curr. Opin. Clin. Nutr. Metab. Care. 10, 136–141. (
- 1990) Delta 6 desaturase in brain and liver during development and aging. Lipids 25, 354–365. , and (
- 2002) Efficiency of conversion of α-linolenic acid to long chain n-3 fatty acids in man. Curr. Opin. Clin. Nutr. Metab. Care. 5, 127–132. (
- 1993) Optimized survival of hippocampal neurons in B27-supplemented Neurobasal, a new serum-free medium combination. J. Neurosci. Res. 35, 567–576. , , and (
- 2007) Docosahexaenoic acid alters bilayer elastic properties. Proc. Natl Acad. Sci. USA 104, 9638–9643. , and (
- 2006) Metabolism of α-linolenic acid in humans. Prostaglandins Leukot. Essent. Fatty Acids 75, 161–168. (
- 2004) Docosahexaenoic acid promotes neurite growth in hippocampal neurons. J. Neurochem. 90, 979–988. and (
- 1999) Reduced palmitate turnover in brain phospholipids of pentobarbital-anesthetized rats. Neurochem. Res. 24, 833–841. , , , and (
- 1999) Carbon recycling into de novo lipogenesis is a major pathway in neonatal metabolism of linoleate and α-linolenate. Prostaglandins Leukot. Essent. Fatty Acids 60, 387–392. , , , and (
- 2005) α-Linolenic acid does not contribute appreciably to docosahexaenoic acid within brain phospholipids of adult rats fed a diet enriched in docosahexaenoic acid. J. Neurochem. 94, 1063–1076. , , , and (
- 2006) Brain elongation of linoleic acid is a negligible source of the arachidonate in brain phospholipids of adult rats. Biochim. Biophys. Acta 1761, 1050–1059. , , , , , and (
- 1976) Biosynthesis of polyunsaturated fatty acids in the developing brain: I. Metabolic transformations of intracranially administered 1-14C linolenic acid. Lipids 11, 67–71. and (
- 2005) The influence of long chain polyunsaturated supplementation on docosahexaenoic acid and arachidonic acid in baboon neonate central nervous system. BMC Med. 3, 11. , , , , and (
- 2001) Essential polyunsaturated fatty acids and the barrier to the brain. The components of a model for transport. J. Mol. Neurosci. 16, 181–193. (
- 1998) Fatty acid transport and utilization for the developing brain. J. Neurochem. 70, 1227–1234. , , , and (
- 2003) Polyunsaturated docosahexaenoic vs docosapentaenoic acid – differences in lipid matrix properties from the loss of one double bond. J. Am. Chem. Soc. 125, 6409–6421. , , , and (
- 2002) Formation of highly reactive A-ring and J-ring isoprostane-like compounds (A4/J4-neuroprostanes) in vivo from docosahexaenoic acid. J. Biol. Chem. 277, 36076–36084. , , et al. (
- 2006) Docosahexaenoic acid prevents neuronal apoptosis induced by soluble amyloid-beta oligomers. J. Neurochem. 96, 385–395. , , et al. (
- 2003) The structure of DHA in phospholipid membranes. Lipids 38, 445–452. , and (
- 2003) Isolation and characterization of unsaturated fatty acids as natural ligands for the retinoid-X receptor. Arch. Biochem. Biophys. 420, 185–193. , , , and (
- 1995) Specific activity of brain palmitoyl-CoA pool provides rates of incorporation of palmitate in brain phospholipids in awake rats. J. Neurochem. 65, 2290–2298. , , , , and (
- 1993) Elongation, desaturation, and esterification of essential fatty acids by fetal rat brain in vivo. J. Lipid Res. 34, 2099–2107. and (
- 2007) Upregulated liver conversion of α-linolenic acid to docosahexaenoic acid in rats on a 15 week n-3 PUFA-deficient diet. J. Lipid Res. 48, 152–164. , , , , and (
- 2007) Estradiol targets synaptic proteins to induce glutaminergic synapse formation in cultured hippocampal neurons: critical role of estrogen receptor-α. J. Neurosci. 27, 6903–6913. , , , and (
- 2007) Docosahexaenoic acid and arachidonic acid are fundamental supplements for the induction of neuronal differentiation. J. Lipid Res. 48, 513–517. , , and (
- 2006) Docosahexaenoic acid promotes neurogenesis in vitro and in vivo. Neuroscience 139, 991–997. , and (
- 2007) β-Amyloid disrupted synaptic vesicle endocytosis in cultured hippocampal neurons. Neuroscience 147, 60–70. and (
- 2000) Inhibition of neuronal apoptosis by docosahexaenoic acid (22:6n-3). Role of phosphatidylserine in antiapoptotic effect. J. Biol. Chem. 275, 35215–35223. , , and (
- 2003) The presynaptic release apparatus is functional in the absence of dendritic contact and highly mobile within isolated axons. Neuron 40, 945–957. , and (
- 2001) The essentiality of long chain n-3 fatty acids in relation to development and function of the brain and retina. Prog. Lipid Res. 40, 1–94. , , and (
- 2005) Where does the developing brain obtain its docosahexaenoic acid? Relative contributions of dietary α-linolenic acid, docosahexaenoic acid, and body stores in the developing rat Pediatr. Res. 57, 157–165. , , and (
- 2003) Disruption of the NMDA receptor-PSD-95 interaction in hippocampal neurons with no obvious physiological short-term effects. Neuropharmacology 45, 738–754. , , and (
- 2005a) N-3 fatty acid deficiency induced by a modified artificial rearing method leads to poorer performance in spatial learning tasks. Pediatr. Res. 58, 741–748. , , and (
- 2005b) An extraordinary degree of structural specificity is required in neural phospholipids for optimal brain function: n-6 docosapentaenoic acid substitution for docosahexaenoic acid leads to a loss in spatial task performance. J. Neurochem. 95, 848–857. , and (
- 2001) Role of docosahexaenoic acid containing phospholipids in modulating G protein-coupled signaling pathways. J. Mol. Neurosci. 16, 153–158. , , and (
- 2005) A role for docosahexaenoic acid-derived neuroprotectin D1 in neural cell survival and Alzheimer disease. J. Clin. Invest. 115, 2774–2783. , , , , , , and (
- 2000) Docosahexaenoic acid, a ligand for the retinoid X receptor. Science 290, 2140–2144. , , , , , and (
- 1990) Role of the blood-brain barrier in the formation of long-chain ω-3 and ω-6 fatty acids from essential fatty acid precursors. J. Neurochem. 55, 391–402. , and (
- 1991) Astrocytes, not neurons, produce docosahexaenoic acid (22:6ω-3) and arachidonic acid (20:4ω-6). J. Neurochem. 56, 518–524. , , , and (
- 1992) The origin of palmitic acid in brain of the developing rat. Lipids, 27, 587–592. , , , and (
- 2007) Endocannabinoid signaling triggered by NMDA receptor-mediated calcium entry into hippocampal neurons. J. Physiol. 584 (2), 407–418. , , , , and (
- 2005) Maternal dietary 22: 6n-3 is more effective than 18: 3n-3 in increasing the 22: 6n-3 content in phospholipids of glial cells from neonatal brain. Brit. J. Nutr. 93, 601–611. , and (
- 2001a) In vivo fatty acid incorporation into brain phospholipids in relation to plasma availability, signal transduction, and membrane remodeling. J. Mol. Neurosci. 16, 243–261. (
- 2001b) Delivery and turnover of plasma-derived essential PUFAs in mammalian brain. J. Lipid Res. 42, 678–685. , and (
- 2007) Polyunsaturated fatty acids stimulate phosphatidylcholine synthesis in PC12 cells. Biochim. Biophys. Acta 1771, 558–563. and (
- 1998) Formation of isoprostane-like compounds (neuroprostanes) in vivo from docosahexaenoic acid. J. Biol. Chem. 273, 13605–13612. , , , , , , and (
- 1996) Arachidonic and docosahexaenoic acids are biosynthesized from their 18-carbon precursors in human infants. Proc. Natl Acad. Sci. USA 93, 49–54. , , and (
- 2001) Mechanism of action of docosahexaenoic acid in the nervous system. Lipids 36, 945–959. , , and (
- 1989) Membrane docosahexaenoate is supplied to the developing brain and retina by the liver. Proc. Natl Acad. Sci. USA 86, 2903–2907. and (
- 2006) Anti-inflammatory actions of neuroprotectin D1/protectin D1 and its natural stereoisomers: assignments of dihydroxy-containing docosatrienes. J. Immunol. 176, 1848–1859. , , , , , , , and (
- 1999) Essentiality of fatty acids. Lipids 34(Suppl.), S1–S3. (
- 1993) Role of cerebrovascular endothelium and astrocytes in supplying docosahexaenoic acid to the brain, in Essential Fatty Acids and Eicosanoids (SinclairA. and GibsonR., eds), pp. 100–103. American Oil Chemist Society, Champaign, IL. and (
- 1995) Reevaluation of the pathways for the biosynthesis of polyunsaturated fatty acids. J. Lipid Res. 36, 2471–2477. , , and (
- 1984) Changes in serum influence the fatty acid composition of established cell lines. In Vitro 20, 732–738. and (
- 2000) Fetal baboons convert 18:3n-3 to 22:6n-3in vivo: a stable isotope tracer study. J. Lipid Res. 42, 581–586. , , , and (
- 1998) γ-Linolenic acid (GLA) is cytotoxic to 36B10 malignant rat astrocytoma cells but not to ‘normal’ rat astrocytes. Brit. J. Cancer. 77, 1612–1620. , , , and (
- 2007) Chronic interleukin-6 alters the level of synaptic proteins in hippocampus in culture and in vivo. Eur. J. Neurosci. 25, 3605–3616. , , , and (
- 2001a) Docosahexaenoic acid synthesis from n-3 polyunsaturated fatty acids in differentiated rat brain astrocytes. J. Lipid Res. 42, 1368–1376. , , , , , and (
- 2001b) Identification of a fatty acid Δ6-desaturase deficiency in human skin fibroblasts. J. Lipid Res. 42, 501–508. , , , , , , and (
- 1996) Is docosahexaenoic acid necessary in infant formula? Evaluation of high linolenate diets in the neonatal rat Pediatr. Res. 40, 687–694. , and (
- 2006) Versatile roles of docosahexaenoic acid in the prenatal brain: from pro- and anti-oxidant features to regulation of gene expression. Prostaglandins Leukot. Essent. Fatty Acids 75, 203–211. (