|
Figure 1. Characterization of the effect of docosahexaenoic acid (DHA) on EAAT2 current. A, a1, representative traces of L-glutamate (L-Glu, 50 μM for 2 min, black bar)-induced current obtained from Xenopus oocytes over-expressing EAAT2 clamped at â50 mV in the absence or presence of DHA (100 μM for 2 min, gray bar). When co-applied, DHA increased the L-Glu-induced EAAT2 current, and the effect disappeared after washout. The structure of DHA is also shown. a2, effect of DHA (100 μM) on the L-Glu concentrationâresponse curves of EAAT2 current at â50 mV. DHA caused a significant decrease of Km for L-Glu, without significantly affecting the mean maximal current, Imax. Km is the Michaelis constant, which is L-Glu concentration needed to achieve a half-maximum binding at equilibrium. a3, representative traces of L-Glu (50 μM for 2 min, black bars)-induced EAAT1 current in the absence or presence of DHA (100 μM for 2 min, gray bars). When the compounds were co-applied, DHA tended to decrease EAAT1 current, and the effect disappeared after washout. Effects of DHA (100 μM) on the L-Glu-induced current amplitudes of EAAT1. The amplitudes were normalized to those just before the application of DHA. B, b1, concentration-dependence of the effect of DHA on L-Glu induced EAAT2 current. b2, currentâvoltage relationship for L-Glu-induced EAAT2 current in the absence or presence of DHA (100 μM). The effect of DHA was independent of holding potential. b3, in the presence of dihydrokainic acid (DHK, 100 μM), a selective inhibitor of EAAT2, DHA no longer enhanced the L-Glu-induced EAAT2 current. C, c1, comparison of the effects of DHA and that of DHA analog. The conjugate of DHA with coenzyme A (DHA-CoA, 100 μM), membrane-impermeable analog of DHA, increased the EAAT2 current to almost the same extent as DHA. DHA methyl ester (DHA-ME, 200 μM), an uncharged analog of DHA, had no effect on L-Glu-induced EAAT2 current. c2, the effect of DHA disappeared when the pH was changed from 7.5 to 5.5. c3, the structures of DHA and the analogs used in these experiments. Error bars represent meanâ±âSD. The numbers written within parentheses in each Figure represent the number of independent experiments. Statistical differences between groups were determined by two-tailed paired Studentâs t test (a2, b3, c1, c2) (denoted by â). p-values are indicated in each Figure panel. EAAT1, excitatory amino acid transporter subtype 1.
|
|
Figure 2. Leu434 residue in re-entrant hairpin loops HP2a is essential for the augmenting effect of DHA on L-Glu-induced EAAT2 current. A, a1, topology of EAAT1, EAAT2, and EAAT1-EAAT2 hybrid chimeras: EAAT2(EAAT1 TM7b-HP2a), EAAT1(EAAT2 TM7b-HP2a), and EAAT1(EAAT2 connector). a2, the effect of DHA on L-Glu-induced currents of EAAT1, EAAT1(EAAT2 TM7b-HP2a), EAAT1(EAAT2 connector), and EAAT2. Data are shown as rates of increase by DHA. B, b1, amino acid alignment from TM7b to HP2a of EAAT2 and EAAT1. The common amino acids are shown on a black background. Single amino acid back mutations were performed at the sites indicated by black arrowheads in EAAT1(EAAT2 TM7b-HP2a) chimera. b2, comparison of the effects of DHA on EAAT1(EAAT2 TM7b-HP2a) chimera and a series of EAAT1(EAAT2 TM7b-HP2a)s with point back mutations to the original amino acid of EAAT1 for the six amino acids shown in b1. Only EAAT1(EAAT2 TM7b-HP2a) L434A shows complete loss of the augmenting effect of DHA. Data are shown as rates of increase by DHA. Exact p-values were 0.7809 for V407L, 0.1898 for M415V, 0.1575 for V426I, 0.4775 for V428I, 0.3042 for L430I, and 0.0006 for L434A. versus EAAT1(EAAT2 TM7b-HP2a group). b3, (top) topology of EAAT2 L434A; (bottom) comparison of the effects of DHA on EAAT2 and EAAT2 L434A. Data are shown as rates of increase by DHA. b4, (top) topology of EAAT1 A435L; (bottom) comparison of the effects of DHA on EAAT1 and EAAT1 A435L. Data are shown as rates of increase by DHA. Error bars represent meanâ±âSD. The numbers written within parentheses in each Figure represent the number of independent experiments. Statistical differences between groups were determined by two-tailed unpaired Studentâs t test (b3 and b4), and Tukeyâs test following one-way factorial ANOVA (a2 and b2) (denoted by â). p-values are indicated in each Figure panel. DHA, docosahexaenoic acid; EAAT1, excitatory amino acid transporter subtype 1; EAAT2, excitatory amino acid transporter subtype 2.
|
|
Figure 3. Identification of other PUFAs that augment L-Glu-induced EAAT2 current. A, structures of fatty acids used in this experiment. ALA, α-linolenic acid; ARA, arachidonic acid; DHA, docosahexaenoic acid; DPA, docosapentaenoic acid, DTetA, docosatetraenoic acid, DTriA, docosatrienoic acid, EPA, eicosapentaenoic acid; OLE, oleic acid. B, b1, effects of fatty acids (100 μM) shown in A on the L-Glu-induced EAAT2 current. DHA, DPA, EPA, ARA, and ALA significantly increased the current. b2, loss of the augmenting effects of DPA, EPA, ARA and ALA in EAAT2 L434A. C, c1, effects of DHA, DPA, EPA, ARA, and ALA (100 μM) on the L-Glu-induced EAAT1 current. DHA significantly decreased the current, while DPA, EPA, ARA, and ALA tended to decrease the current. c2, the effects of DPA, EPA, ARA, and ALA on L-Glu-induced EAAT1 A435L current. These PUFAs augmented the L-Glu-induced EAAT1 A435L current. Error bars represent meanâ±âSD. The numbers written within parentheses in each Figure represent the number of independent experiments. Statistical differences between groups were determined by two-tailed unpaired Studentâs t test (b2 and c2), and Tukeyâs test following one-way factorial ANOVA (b1 and c1 versus OLE-treated group) (denoted by â). p-values are indicated in each Figure panel. EAAT2, excitatory amino acid transporter subtype 2; PUFA, polyunsaturated fatty acid.
|
|
Figure 4. Proposed binding conformation for DHA in the transport/trimerization domain interface of EAAT2 homology model in the outward facing state OFS. A, a1, extracellular view of trimerized EAAT2 OFS homology model based on EAAT1 crystal structure. Trimerization domain is shown in green ribbon. Transporter domain is shown in gray surface. Structural data were presented using graphical user interface in Maestro Suite. The homology model of EAAT2 was constructed as a monomer based on the crystal structure of OFS EAAT1 (PDBID: 5LLM) with energy-based loop refinement using Homology Modeling unit in Maestro Suite. The quality of homology model was checked by PROCHECK. a2 and a3, magnified monomer in the hatched square in a1 in the absence (a2) or presence (a3) of DHA. The lipid crevice calculated by SiteMap exists at the interface between trimerization domains and transport domains (yellow space) (a2). DHA is docked to the lipid crevice (carbon: purple spheres; hydrogen: white spheres) (a3). B, b1 and b2, docking poses of DHA to the lipid pocket in the vicinity of HP2 according to induced fit docking protocol. The trimerization domain and transport domain are shown in green and gray ribbons, respectively. Carbons in DHA and EAAT2 L434 are represented by purple and yellow sticks, respectively. The atoms in L-Glu are shown as follows: carbon: blue sphere; hydrogen: white sphere, oxygen: red sphere; nitrogen: hiding. Na+ is shown as a pink sphere. Two types of the DHA conformations could be visualized according to the position of the carboxylic group, i.e., one is with carboxyl group on upper side (b1) and the other is with carboxyl group on lower side (b2). Both of them have similar U-shaped conformation. Inset is the DHA conformations in each case. Three-dimensional position of DHA is in close proximity to L-Glu binding site and Na+ binding site. DHA, docosahexaenoic acid; EAAT2, excitatory amino acid transporter subtype 2; OFS, outward facing state.
|
