Vitamin E functions by association with a novel binding site on the 67 kDa laminin receptor activating diacylglycerol kinase

Abstract

It is generally recognized that the main function of α-tocopherol (αToc), which is the most active form of vitamin E, is its antioxidant effect, while non-antioxidant effects have also been reported. We previously found that αToc ameliorates diabetic nephropathy via diacylglycerol kinase alpha (DGKα) activation in vivo, and the activation was not related to the antioxidant effect. However, the underlying mechanism of how αToc activates DGKα have been enigmatic. We report that the membrane-bound 67 kDa laminin receptor (67LR), which has previously been shown to serve as a receptor for epigallocatechin gallate (EGCG), also contains a novel binding site for vitamin E, and its association with Vitamin E mediates DGKα activation by αToc. We employed hydrogen-deuterium exchange mass spectrometry (HDX/MS) and molecular dynamics (MD) simulations to identify the specific binding site of αToc on the 67LR and discovered the conformation of the specific hydrophobic pocket that accommodates αToc. Also, HDX/MS and MD simulations demonstrated the detailed binding of EGCG to a water-exposed hydrophilic site on 67LR, while in contrast αToc binds to a distinct hydrophobic site. We demonstrated that 67LR triggers an important signaling pathway mediating non-antioxidant effects of αToc, such as DGKα activation. This is the first evidence demonstrating a membrane receptor for αToc and one of the underlying mechanisms of a non-antioxidant function for αToc.

Keywords: 67 kDa laminin receptor; diabetic nephrophaty; diacylglycerol kinase; epigallocatechin gallate; vitamin E.

Fig. 1.. Translocation of DGKα to the plasma membrane induced by tocopherols.

(a) Structure of αToc. (b) Localization of GFP-DGKα in DDT1-MF-2 cells before and 3 minutes after 200 μM αToc stimulation. (c) The translocation rate of GFP-DGKα in DDT1-MF-2 cells by 200 μM αToc in the absence or presence of anti-67LR antibody treatment. N=4. (d) Effect of the knockdown of 67LR on the translocation of GFP-DGKα in DDT1-MF-2 cells by 200 μM αToc. N=3. (e) Localization of GFP-DGKα in DDT1-MF-2 cells before and 3 minutes after 200 μM tocopherol stimulation. The difference in the position of the methyl moiety in the structure of each of the tocopherols is shown above the respective pictures. (f) The translocation rate of GFP-DGKα in DDT1-MF-2 cells induced by each of the tocopherols. N=3. The data shows the percentage of the translocation rate relative to αToc. The values are mean ±SE. Asterisks indicate significance (**p < 0.01). Full-length blots are presented in Fig. S1.

Fig. 2.. Direct binding between αToc and 67LR and its affinity.

(a) The result of the binding assay between 10 μM αToc and precipitated FLAG-67LR or FLAG-α-TTP using HPLC is shown. The expression and precipitation of FLAG-67LR and FLAG-α-TTP were detected by western blot. The value is the ratio of bound αToc to FLAG alone. N=4. (b) Concentration-dependent binding of αToc to 67LR detected by the binding assay using HPLC. The value was shown as a percentage of 10 μM. (c) The result of the binding assay using tocopherol conjugated beads. The amount of precipitated FLAG-67LR by the control and tocopherol conjugated beads was detected by western blot. (d) The representative raw data of ITC for titration of 120 μM GST-67LR-His to 20 μM αToc at 298K. (e) The integrated area under the curve of each peak from (d) was plotted by the molar ratio of GST-67LR-His and αToc. The best fit curve was drawn by the Origin software, and the K_d value was calculated based on the curve. The intensity of bands in the western blots were analyzed by Image J. The error bars show SE. Asterisks indicate significance (*p < 0.05). Full-length blots are presented in Fig. S2 and S4.

Fig. 3.. The putative αToc binding site on 67LR determined by HDX/MS.

The effect of overexpression of FLAG-67LR Lys166Glu on 200 μM (a) EGCG or (b) αToc-induced translocation of GFP-DGKα in DDT1-MF-2 cells. N=3 (c) The image of the EGCG binding site of 67LR (green: hydrophilic surface). The docking simulations of (d) EGCG and (e) αToc to the putative EGCG binding site of 67LR by Glide software. Hydrogen bonds are shown as a yellow dashed line. (f) The change in the number of deuterium in the typical reproduced regions by αToc (Cyan: Lys11-Lys17, Blue: Asn50-Ala61, Orange: Trp176-Leu183). (g) The change in the number of deuterium in the typical reproduced regions by EGCG (Red: Leu16-Met34, Yellow: Ile161-Gly172). The dashed lines point the residues neighboring color-coded residues show no significant difference in number of deuterium. The amino acid sequence includes the remaining N-terminal linker to the GST tag and the C-terminal 6×His tag. The error bars show SE. Asterisks indicate significance (*p < 0.05).

Fig. 4.. Binding simulations of αToc and EGCG to their respective putative binding sites.

The binding mode of (a) αToc and (c) EGCG in 67LR (the final frame of the simulations shown in Movie S1 and S3). The typical amino acid residues in the binding mode of (b) αToc and (d) EGCG in 67LR. The structure of αToc and EGCG are shown in black, and the colored regions are the same as shown in Fig. 3, respectively. The surface of the protein was shown in a gray transparent cloud. The hydrogen bonds are shown as a yellow dashed line (distance cutoff: 3.2Å, angle cutoff: 30°).

Fig. 5.. The effect of mutants in the αToc binding site on the translocation of DGKα.

(a) The position of mutated amino acids in the αToc binding region is shown. The effect of overexpression of the mutants on the 200 μM (b) αToc (N=3) or (c) EGCG (N=3) -induced translocation of GFP-DGKα in DDT1-MF-2 cells. The Control is the overexpressed FLAG vector alone in the cells. The error bars show SE. Asterisks indicate significance comparing to Control (*p < 0.05, **p < 0.01).

Fig. 6.. Effect of the knockdown of 67LR on the anticancer effect of EGCG and αToc.

The effects of various concentrations of (a) EGCG (N=3) or (b) αToc (N=3) on the viability of HCT116 cells. The value is shown as a percentage of that seen with no addition (0 μM). (c) Changes in the expression of 67LR after siRNA transfection. The effects of the knockdown of 67LR on the (d) 50 μM EGCG (N=3) or (e) 100 μM αToc (N=3) -induced decrease of HCT116 cell viability. The error bars show SE. Asterisks indicate significance comparing to respective control (*p < 0.05).

Fig. 7. Model for the EGCG and αToc binding sites on 67LR and its functions.

The transparent purple surface shows the entire 67LR. The green surface shows a hydrophilic EGCG binding site facing the outer environment. The orange surface shows a hydrophobic αToc binding site embedded in the membrane.