Identification and Characterization of Synaptic Vesicle Membrane Protein VAT-1 Homolog as a New Catechin-Binding Protein

Abstract

(−)-Epigallocatechin-3-gallate (EGCg), a major constituent of green tea extract, is well-known to exhibit many beneficial actions for human health by interacting with numerous proteins. In this study we identified synaptic vesicle membrane protein VAT-1 homolog (VAT1) as a novel EGCg-binding protein in human neuroglioma cell extracts using a magnetic pull-down assay and LC–tandem mass spectrometry. We prepared recombinant human VAT1 and analyzed its direct binding to EGCg and its alkylated derivatives using surface plasmon resonance. For EGCg and the derivative NUP-15, we measured an association constant of 0.02–0.85 ×10³ M⁻¹s⁻¹ and a dissociation constant of nearly 8 × 10⁻⁴ s⁻¹. The affinity K_m(affinity) of their binding to VAT1 was in the 10–20 µM range and comparable with that of other EGCg-binding proteins reported previously. Based on the common structure of the compounds, VAT1 appeared to recognize a catechol or pyrogallol moiety around the B-, C- and G-rings of EGCg. Next, we examined whether VAT1 mediates the effects of EGCg and NUP-15 on expression of neprilysin (NEP). Treatments of mock cells with these compounds upregulated NEP, as observed previously, whereas no effect was observed in the VAT1-overexpressing cells, indicating that VAT1 prevented the effects of EGCg or NUP-15 by binding to and inactivating them in the cells overexpressing VAT1. Further investigation is required to determine the biological significance of the VAT1–EGCg interaction.

INTRODUCTION

A variety of plants produce natural polyphenols as secondary products, which show many beneficial actions for human health, such as anti-oxidation, anti-atherosclerosis, anti-inflammation, antibiosis, vasodilatation, neuroprotection, and anti-aggregation activity against misfolding proteins, as demonstrated using various experimental paradigms in vitro and in vivo.^1,2) It is often observed that certain polyphenols, like (−)-epigallocatechin-3-gallate (EGCg), have many of the actions mentioned above. For example, dietary intake of green tea, which has a large amount of EGCg, contributes to maintenance of human health^3,4); however, it is difficult to use multifunctional compounds as pharmaceutical agents. In contrast, although the intracellular signaling pathways of polyphenol-induced up- or down-regulation of genes have been well documented,⁴⁾ it is still unclear whether polyphenols interact directly with cell-surface receptors or intracellular binding proteins. In the case of EGCg, some molecules have already been identified as EGCg-binding proteins.⁵⁾

Because the hydrophilicity of natural polyphenols results in poor bioavailability, trials to increase their hydrophobicity by adding alkylated chains or rings have been carried out in some laboratories, including ours.^6–9) Such structural modification of natural polyphenols should not only result in the improvement of their bioavailability but also specialization or potentiation of certain functions of multifunctional polyphenols, as has been shown in traditional medicinal chemistry of antibiotics. Previously, we used a polyphenol library to screen for compounds that could enhance the activity and expression of neprilysin (NEP), the major enzyme degrading the amyloid-β peptide (Aβ), aggregates of which are the hallmark of Alzheimer’s disease. We found that EGCg, amentoflavone, apigenin, kaempferol, and chrysin upregulated NEP in a cell culture system.⁹⁾ Interestingly, we successfully converted an ineffective natural compound such as quercetin into an effective one by adding an aliphatic moiety.⁹⁾ Thus, structural modification enables us to convert an inactive compound to active one.

We hypothesized that polyphenols, such as EGCg, may enhance the activity and expression of NEP by interacting with any proteins in the cells. Therefore, in the present study, we isolated EGCg-binding proteins in human neuroglioma (H4) cell extracts using a pull-down assay with EGCg-coupled magnetic beads and identified synaptic vesicle membrane VAT-1 homolog (VAT1; UniProtKB/Swiss-Prot Q99536-VAT1_HUMAN) as a catechin-binding protein using mass spectrometry. Then, we prepared recombinant VAT1, measured its ability to bind EGCg using surface plasmon resonance spectroscopy and predicted a possible binding moiety of EGCg with VAT1 using some EGCg derivatives with an aliphatic group at different sites. Lastly, we investigated the biological activity of VAT1, which may be affected by its interaction with catechins.

MATERIALS AND METHODS

Polyphenols

EGCg and aliphatic EGCg derivatives were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan), from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan), or prepared in a previous study,⁸⁾ where Fudouji compounds were renamed NUP compounds (e.g., Fudouji-15 was renamed NUP-15) following further optimization of their chemical structure and construction of a compound library for drug development (Fig. 1). The polyphenols were dissolved in dimethyl sulfoxide (DMSO; Nacalai Tesque Inc., Kyoto, Japan) and stored at −20 °C as 20 mM stock solutions until use. Other chemical reagents were of a commercially available grade.

Fig. 1. Chemical Structures of (−)-Epigallocatechin-3-gallate (EGCg) and Its Derivatives Used in This Study

Cell Culture and Preparation of Cell Lysate

H4 cells overexpressing human amyloid precursor protein with the Swedish mutation (H4-NL cells)¹⁰⁾ were cultured in Dulbecco’s modified Eagle’s medium (Nacalai Tesque Inc.) supplemented with 10% (v/v) fetal bovine serum (Merck KGaA, Darmstadt, Germany), 100 U/mL penicillin, and 100 µg/mL streptomycin (Nacalai Tesque Inc.) at 37 °C under 5% (v/v) CO₂. H4-NL cells were seeded on 6-well culture plates at a density of approximately 1.2 × 10⁵ cells per well. After 24 h, the culture medium was replaced with Opti-MEM I reduced-serum medium (Thermo Fisher Scientific K.K., Tokyo, Japan) containing 10 µM polyphenol (final DMSO concentration: 0.1% (v/v)).

Cells treated with polyphenols for 48 h were harvested and lysed in 40 µL lysis buffer A (50 mM Tris–HCl pH 7.4, 0.15 M NaCl, 1% (v/v) Triton X-100 [Nacalai Tesque Inc.], protease inhibitor cocktail [cOmplete ethylenediaminetetraacetic acid (EDTA)-free; Roche Diagnostics GmbH, Mannheim, Germany], 10 µM Z-Leu-Leu-Leu-CHO [Peptide Institute Inc., Osaka, Japan], and 0.7 µg/mL pepstatin A [Peptide Institute Inc.]) with a pellet mixer and then stood on ice for 1 h. The obtained cell lysates were centrifuged at 21000 × g and 4 °C for 30 min, and the supernatants were collected for measurement of NEP activity and Western blot analysis. Protein concentrations were determined using a BCA Protein Assay Kit (TaKaRa Bio Inc., Shiga, Japan).

Coupling of EGCg to Magnetic Beads

To prepare catechin-coupled magnetic beads (catechin beads), we chemically synthesized EGCg derivatives by reacting the alkyl-chain spacer 15-hydroxypentadecanoic acid with EGCg. As a result, we obtained two derivatives: NUP-E15-1 (ligand #2), with alkyl chains bound to either the upper or lower side of the C-ring of EGCg via tetrahydrofuran; and NUP-E15-2 (ligand #3), with alkyl chains bound to both the upper and lower side of the C-ring of EGCg via tetrahydrofuran. In both NUP-E15-1 and NUP-E15-2, a catechol or pyrogallol moiety of B- and G-rings of EGCg is preserved (Supplementary Figs. 1 and 2).

To synthesize NUP-E15-1 and NUP-E15-2, EGCg (460 mg) was dissolved in acrolein (2.0 mL) and heated at 70 °C for 1 h. The mixture was applied to a column of Sephadex LH-20 (2.0 cm i.d. × 15 cm) and eluted first with EtOH (100 mL) and then an EtOH/acetone/H₂O mixture (8 : 1 : 1 v/v/v, 200 mL) to give EGCg-Acr (445 mg). EGCg-Acr (114 mg, 0.2 mmol) and 15-hydroxypentadecanoic acid (52 mg, 0.2 mmol) were dissolved in acetone (2.0 mL) containing trifluoroacetic acid (5% (v/v)) in a screw-capped vial, and the mixture was heated at 60 °C for 7 h. The products were separated using silica gel column chromatography (2.0 cm i.d. × 10 cm) with stepwise CHCl₃/MeOH elution (1 : 0, 9 : 1, 85 : 15, 8 : 2, 75 : 25, and 7 : 3 v/v, each 100 mL) to yield a mixture of mono-substituted NUP-E15-1a and -1b products (46.8 mg) and a disubstituted NUP-E15-2 product (18.5 mg).

EGCg-Acr was obtained as an off-white amorphous powder. [α]_D −146.8 (c = 0.10, MeOH); FAB-MS m/z: 571 [M + H]⁺, high-resolution (HR)-FAB-MS m/z: 571.1450 [M + H]⁺ (C₂₈H₂₇O₁₃ calcd 571.1452);. UV λ_max (MeOH) nm (log ε): 210 (3.87), 275 (2.98); IR ν_max cm⁻¹: 3374, 2954, 2856, 1688, 1613, 1534, 1453; ¹H-NMR (400 MHz, acetone-d₆) δ: 5.05 (1H, br s, H-2), 5.54 (1H, br s, H-3), 2.84 (1H, br d, H-4), 3.00 (m, H-4), 6.56 (2H, br s, B-ring H-2,6), 7.00 (2H, br s, galloyl H-2,6), 5.54 (2H, m, Acr-1,1′), 1.87 (4H, m, Acr-2,2′), 2.56, 2.7 5(m, Acr-3,3′); ¹³C-NMR (125 MHz, acetone-d₆) δ: 77.7 (C-2), 69.1 (C-3), 28.0 (C-4), 99.6 (C-4a), 151.2 (C-5), 103.3 (C-6), 151.2 (C-7), 102.6 (C-8), 149.7 (C-8a), 130.8 (B-ring C-1), 106.4 (B-ring C-2,6), 146.2 (B-ring C-3,5), 133.0 (B-ring C-4), 121.8 (galloyl C-1), 109.7 (galloyl C-2,6), 145.8 (galloyl C-3,5), 138.7 (galloyl C-4), 166.2 (galloyl C-7). 92.9 (Acr-C-1,1′), 26.7 (Acr-C-2,2′), 15.9 (Acr-C-3,3′).

NUP-E15-1a and -1b were obtained as a white amorphous powder. [α]_D −102.7 (c = 0.10, MeOH); FAB-MS m/z: 849 [M + K]⁺, HR-FAB-MS m/z: 849.3092 [M + K]⁺ (C₄₃H₅₄O₁₅K calcd 849.3100). UV λ_max (MeOH) nm (log ε): 209 (3.89), 275 (3.01); IR ν_max cm⁻¹: 3250, 2923, 2851, 1682, 1615, 1536, 1455; ¹H-NMR (400 MHz, acetone-d₆) δ: 5.06 (H-2), 5.20, 5.28 (H-3), 2.96 (H-4), 6.66 (br s, B-ring H-2,6), 6.99 (br s, galloyl C-2,6), 5.54 (m, Acr-H-1,1′), 1.90 (m, Acr-H-2,2′), 2.60 (m, Acr-H-3,3′), 1.2–1.6 (m, CH₂), 3.55, 3.88 (CH₂-O-); ¹³C-NMR (125 MHz, acetone-d₆) δ: 77.6 (C-2), 68.5–69.3 (C-3), 28.0 (C-4), 100.0 (C-4a), 151.2 (C-5), 103.0–103.7 (C-6,8), 151.2 (C-7), 149.5 (C-8a), 130.9 (B-ring C-1), 106.4 (B-ring C-2,6), 146.2 (B-ring C-3,5), 132.9 (B-ring C-4), 121.7 (galloyl C-1), 110.0 (galloyl C-2,6), 145.8 (galloyl C-3,5), 138.8 (galloyl C-4), 166.2 (galloyl C-7), 92.8, 97.9 (Acr-C-1,1′), 26.8 (Acr-C-2,2′), 15.9 (Acr-C-3,3′), 175.0 (COOH), 25.6, 26.7, 29.2, 29.6, 29.8, 30.0, 30.2, 30.4, 34.2 (CH₂), 68.5–69.3 (CH₂O).

NUP-E15-2 was obtained as a white amorphous powder. [α]_D −86.6 (c = 0.10, MeOH); FAB-MS, m/z: 1089 [M + K]⁺, HR-FAB-MS m/z: 1089.5195 [M + K]⁺ (C₅₈H₈₂O₁₇K requires 1089.5189); UV λ_max (MeOH) nm (log ε): 209 (3.77), 275 (3.05); IR ν_max cm⁻¹: 3330, 2924, 2852, 1703, 1615, 1536, 1459; ¹H-NMR (400 MHz, acetone-d₆) δ: 5.04, 5.08 (H-2), 5.20, 5.26 (H-3), 2.90–3.10 (H-4), 6.66 (br s, B-ring H-2,6), 6.98, 7.01 (galloyl C-2,6), 5.55 (m, Acr-H-1,1′), 1.90 (m, Acr-H-2,2′), 2.50–2.85 (m, Acr-H-3,3′), 1.2–1.6 (m, CH₂), 3.50–3.85 (CH₂-O-). ¹³C-NMR (125 MHz, acetone-d₆) δ: 77.6 (C-2), 68.2–69.3 (C-3), 26.8 (C-4), 100.1 (C-4a), 151.2 (C-5), 103.2–103.8 (C-6,8), 151.2 (C-7), 149.2 (C-8a), 130.9 (B-ring C-1), 106.4 (B-ring C-2,6), 146.3 (B-ring C-3,5), 133.0 (B-ring C-4), 121.7 (galloyl C-1), 110.0 (galloyl C-2,6), 145.8 (galloyl C-3,5), 138.7 (galloyl C-4), 166.2, 166.3 (galloyl C-7). 97.7, 97.9, 98.4 (Acr-C-1,1′), 26.8 (Acr-C-2,2′), 15.5 (Acr-C-3,3′), 174.8 (COOH), 225.6, 26.7, 30.4, 34.2 (CH₂, overlapped with solvent signals), 68.2–69.3 (CH₂O).

Catechin beads were prepared by coupling NUP-E15-1 and NUP-E15-2 to FG beads NH₂ (TAS8848N1130, Tamagawa Seiki Co., Ltd., Nagano, Japan) and Dynabeads M-270 Amine (Thermo Fisher Scientific Inc., Waltham, MA, U.S.A.), following the manufacturers’ protocols (Supplementary Fig. 2). Briefly, 10 mM NUP-E15-1 or NUP-E15-2, 10 mM N-hydroxysulfosuccinimide (H-1304, Tokyo Chemical Industry Co., Ltd., Tokyo, Japan), and 10 mM 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (15022-86, Nacalai Tesque) were mixed in dimethyl formamide for 2 h at room temperature to activate the carboxyl group at the terminus of the alkyl chain of NUP-E15-1 or NUP-E15-2. After the reaction, 1.7 mg of FG NH₂ beads was added to the above mixture and incubated overnight at room temperature. After the incubation, the beads were washed with dimethyl formamide and remaining free amino groups of the beads were masked with acetic anhydride and triethylamine for 2 h. Then the beads were treated with 0.1 M NaOH for 30 min and washed with phosphate-buffered saline (PBS). All steps in the above procedure were carried out by mixing gently with a programmable rotator (Multi Bio RS-24, BIOSAN Ltd., Riga, Latvia). As a negative control, we prepared magnetic beads coupled to 15-hydroxypentadecanoic acid (ligand #1) as described above.

Pull-Down of EGCg-Binding Protein Candidates Using Catechin Beads

H4-NL cells were harvested and homogenized in chilled PBS containing protease inhibitor cocktail cOmplete EDTA-plus (Roche Diagnostics GmbH) in the glass tube of a Potter–Elvehjem tissue homogenizer at 2000 rpm for 20 strokes using an overhead stirrer with a Teflon pestle. The homogenate was ultracentrifuged at 70000 × g using an Optima TLX ultracentrifuge and a TLA110 rotor (Beckman Coulter, Inc. Brea, CA, U.S.A.) for 29 min at 4 °C. The obtained pellet was suspended in the homogenization buffer using a pellet mixer and ultracentrifuged as described above to prevent contamination of cytosolic proteins. The pellet was lysed in 40 µL lysis buffer B (PBS pH 7.4, 1% (v/v) Triton X-100, cOmplete EDTA-plus) using a pellet mixer and then stood on ice for 90 min. The lysate was again ultracentrifuged as described above, and the supernatants were collected as a membrane fraction. Protein concentrations were determined using a BCA Protein Assay Kit (TaKaRa Bio Inc.).

The membrane fraction (500 µg/200 µL) was incubated with the catechin beads at 4 °C overnight by rotating gently with a programmable rotator. After incubation, the catechin beads were washed with phosphate buffer (PB) containing 1% (v/v) Triton X-100 and magnetized, then the wash buffer was discarded. EGCg-binding protein candidates were sequentially eluted at stepwise concentrations of NaCl (0.15, 0.6, and 1.0 M) in 1% (v/v) Triton X-100/PB (pH 7.4), 0.5 M NaCl/20 mM MES (pH 6), and 50 mM citrate buffer (pH 3). The eluates were mixed with concentrated sodium dodecyl sulfate (SDS) sample buffer, and candidate proteins were separated using 10% gradient SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and visualized using a Silver Stain MS kit (299-58901; Fujifilum Wako Pure Chemical Corporation, Osaka, Japan). A portion of gel containing a candidate protein was cut out and analyzed using LC–tandem mass spectrometry (LC-MS/MS).

Mass Spectrometry Analysis

A portion of gel containing a candidate protein was destained using reagents supplied in the Silver Stain MS kit and washed three times using 50 mM ammonium hydrogen carbonate solution. The protein was reduced by reaction with 50 mM ammonium hydrogen carbonate solution containing 10 mM dithiothreitol at 56 °C for 1 h and alkylated by reaction with 50 mM ammonium hydrogen carbonate solution containing 10 mg/mL iodoacetamide at room temperature for 45 min. The reaction mixture was dehydrated using an evaporator (CC-10; TOMY SEIKO Co., Ltd., Tokyo, Japan) and digested with trypsin at the final concentration of 0.012 µg/µL (v5280; Promega Corporation, Madison, WI, U.S.A.) in 50 mM aqueous ammonium hydrogen carbonate containing 10% (v/v) acetonitrile (ACN) for 16 h before undergoing a three-step extraction procedure with sequential 100 µL volumes of 30% (v/v) ACN/0.07% (v/v) trifluoroacetic acid (TFA), 50% (v/v) ACN/0.05% (v/v) TFA, and 80% (v/v) ACN/0.02% (v/v) TFA. The extract was evaporated and dissolved in 50 µL of 5% (v/v) ACN/0.1% (v/v) TFA and desalted using a GL-SDB chip filter (GL Sciences Inc., Tokyo, Japan). Finally, the extract was evaporated again, dissolved in 2% (v/v) ACN/0.1% TFA, and analyzed using a custom nano-LC–electrospray ionization–MS/MS system (LTQ-XL; Thermo Fisher Scientific). MS/MS data were extracted and searched against the human protein database (UniProt) using Proteome Discoverer v.3.3 (Thermo Fisher Scientific).

Construction of Human VAT1 Expression Vectors

Human VAT1 cDNA in pOTB7 (IRAL005E11) was provided by RIKEN BioResource Centre through the National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology, Japan. First, the open reading frame (ORF) of VAT1 cDNA was cloned at the HindIII–XhoI sites of pcDNA3.1(+) (Invitrogen) using a PCR-based cloning method. Then the VAT1 ORF was introduced into the KpnI–XhoI sites of pEBMulti-Neo TARGET tag-N (FUJIFILM Wako Pure Chemical Corporation), which is an episomal vector designed to add a TARGET tag sequence to the N-terminus of a gene of interest, to prepare H4-NL cells overexpressing VAT1 stably. For preparation of human VAT1 recombinant protein, the VAT1 ORF was introduced into the NotI–XhoI sites of the pET28a(+) vector (69864-3CN, Novagen, Merck KGaA) using a PCR-based cloning method with a primer set (forward primer: NotI–His tag–target F, 5′-tataccatgggcagcagccatcatcatcatcatcacgaattcattgagggtcgctaccccg-3′; reverse primer: XhoI–VAT1, 5′-tcactcgagcgctagttctccttctctggc-3′). The nucleotide sequences of VAT1 cDNA and around the ligation sites in the plasmid vectors were confirmed using the DNA sequencing service of Eurofins Genomics, Tokyo, Japan.

Preparation of Histidine (His)-Tagged Human VAT1 Recombinant Protein

We transformed BL21(DE3)pLysS competent cells (D00076000, Novagen, Merck KGaA) with the pET28a-VAT1 plasmid and selected colonies on Luria–Bertani (LB) agar plates with 25 µg/mL kanamycin. The cells were cultured with shaking in 3 mL of LB medium containing 60 µg/mL kanamycin for 4 h at 37 °C, then the cell culture was scaled up to 400 mL of fresh LB medium with 60 µg/mL kanamycin in a 1 L flask and incubated at 37 °C until optical density of the Escherichia coli solution reached 0.6–0.8 at 600 nm. Then isopropyl-β-D-thiogalactopyranoside (19742-81; Nacalai Tesque, Inc.) was added to a final concentration of 0.8 mM and incubated with shaking at 18 °C for 24 h to induce expression of VAT1.

After 24 h incubation, the E. coli solution was centrifuged, and the pellet was collected and stored at −80 °C until use. The cell pellet was suspended in lysis buffer C (50 mM Tris–HCl, 150 mM NaCl, pH 8.0) containing 0.5 mg/mL lysozyme and 1 : 100 v/v Protease Inhibitor Cocktail Set VII DMSO solution (×100, 167-26101, FUJIFILM Wako Pure Chemical Corporation) and kept on ice for 20 min. The cells were lysed by applying eight cycles of ultrasonic disruption on ice (on and off for 10 and 20 s, respectively) using an ultrasonic disruptor (UD-211, LST-100; Tomy Seiko Co., Ltd.), then treated with deoxyribonuclease (DNase) I (final concentration: 20 µg/mL) and kept on ice for 1 h. The cell lysate was centrifuged at 20400 × g (MX-201, Tomy Seiko Co., Ltd.) and 4 °C for 10 min and the supernatant collected. Small aggregates in the supernatant were solubilized in Triton X-100 (final concentration: 1% (v/v)) on ice for 1 h, during which time the supernatant was treated with two cycles of sonication (on and off for 10 and 20 s, respectively) on ice. Finally, the supernatant was centrifuged at 20400 × g and 4 °C for 10 min and then passed through a syringe filter Minisart RC4 (1705001VS, Sartrius Stedium Lab Ltd., Stonehouse, U.K.) to yield the E. coli extract.

Purification of the His-tagged human VAT1 from the E. coli extract was carried out using a stepwise concentration of imidazole and an ÄKTA pure 25 L1 fast protein liquid chromatography system (GE Healthcare Life Sciences, Little Chalfont, Buckinghamshire, U.K.) equipped with a HisTrap HP column (17-5247-01, GE Healthcare Life Sciences). The stepwise concentration of imidazole was formed by mixing buffer A (20 mM Tris–HCl, 300 mM NaCl, 5 mM imidazole, pH 8.0) with buffer B (20 mM Tris–HCl, 300 mM NaCl, 500 mM imidazole, pH 8.0). Purification conditions were as follows: sample application in 100% buffer A, 0.5 mL/min flow rate; wash with 96% buffer A, 4.0% buffer B (24.8 mM imidazole), 1 mL/min flow rate, 60 min duration; stepwise elution with 14, 17, and 50% buffer B (74.3, 89.2, and 252.5 mM imidazole, respectively), 1 mL/min flow rate (13, 13, and 10 min duration, respectively). The eluate was automatically fractionated every 500 µL on a 96-well deep plate and analyzed using SDS-PAGE, Western blotting, and silver staining using the Silver Stain MS kit. The fractions including His-tagged VAT1 protein were collected, then a Vivaspin turbo 4 device (3000 Da cut-off) was used to concentrate them and exchange the elution buffer with buffer C (20 mM Tris–HCl buffer, 300 mM NaCl and 20% (v/v) glycerol). The concentrated protein sample was then again subjected to SDS-PAGE and silver staining to check the purity of recombinant protein.

Surface Plasmon Resonance (SPR) Analysis

The binding affinities of catechins and candidate proteins were measured using a Biacore T200 SPR system (Cytiva, Tokyo, Japan). Anti-His-tag antibody (MBL International, U.S.A.) was diluted in sodium acetate solution (pH 5.0) at a final concentration of 20 µg/mL and immobilized on a CM5 sensor chip (Cytiva) using amine coupling. Then, His-tagged recombinant human VAT1 was captured by the immobilized anti-His-tag antibody at a concentration of 10 µg/mL. The running buffer was HBS-EP+ (0.01 M HEPES, 0.15 M NaCl, 0.003 M EDTA, 0.05% (v/v) polysorbate 20, pH 7.4; Cytiva) containing 0.5% (v/v) DMSO. Six concentrations of catechin (0, 1.25, 2.5, 5, 10, and 20 µM) were injected at a flow rate of 30 µL/min; the contact time was set to 2 min, and the dissociation time was set to 10 min. Data were recorded at 25 °C and analyzed using Biacore T200 Evaluation Software (v3.1, Cytiva). Following the manufacturer’s instructions, we evaluated the reliability of the kinetic rate constants using the uniqueness value (U-value) as an indicator. The U-values obtained in all experiments were ≤15 and considered to be reliable. The association rate constant (k_a) and dissociation rate constant (k_d) were determined from the sensorgrams, and the equilibrium dissociation constant was calculated as K_D = k_d/k_a. The maximal SPR signals (recorded 120 s before changing from compound-containing buffer to compound-free buffer) were analyzed using the Michaelis–Menten equation, and K_D values were calculated using Biacore T200 Evaluation Software.

Western Blot Analysis

The cell lysates were dissolved in 1 × Laemmli sample buffer at a concentration of 0.3 µg/µL, and aliquots containing 3 µg of protein were separated using 7% gradient SDS-PAGE and electro-transferred onto 0.45 µm polyvinylidene difluoride membranes (Immobilon-P; Merck KGaA). The membranes were probed with a goat polyclonal antibody against human NEP (0.2 µg/mL; R&D Systems, Minneapolis, MN, USA; catalogue number [Cat#] AF1182; research resource identifier [RRID] AB_354652), a mouse monoclonal antibody against TARGET-tag (1 : 3,000 dilution; FUJIFILM-Wako Cat# 016-25481), or a rabbit polyclonal antibody against VAT1 (1 : 500 dilution; Sigma-Aldrich Cat# HPA045170; RRID AB_10964166), followed by horseradish peroxidase-conjugated anti-goat, -mouse, or -rabbit immunoglobulin G (IgG) (Cytiva). Immunoreactive bands on the membranes were visualized using an enhanced chemiluminescence kit (Immunostar LD; FUJIFILM Wako Pure Chemical Corporation), and the band intensities were determined using a chemiluminescent imager (LAS4000; Fuji Photo Film, Tokyo, Japan) and Science Lab 97 Image Gauge software (ver. 3.0.1; FUJIFILM Wako Pure Chemical Corporation). The relative immunoreactive protein content in each sample was calculated by reference to a standard curve constructed using certain samples in the control group to minimize variations in the immunoreactivity and intensity of ECL signals among different experiments. The membranes were re-probed using an anti-β-actin antibody (1 : 10,000 dilution; Sigma-Aldrich, St. Louis, MO, USA; Cat# A1978; RRID AB_476692) to confirm that equal amounts of total protein had been loaded on the gel.

Establishment of Cell Lines Stably Expressing VAT1

H4-NL cells were seeded on 12-well culture plates at a density of 7.1 × 10⁴ cells per well. After 24 h, pEBMulti-Neo TARGET-human VAT1 or the empty vector (1.6 µg/100 µL Opti-MEM/well) was transfected into the cells using Lipofectamine 2000 (4 µL/100 µL Opti-MEM/well, Thermo Fisher Scientific Inc.), following the manufacturer’s protocol. After 24 h, cells were released from the plate using a solution including 0.25% (w/v) trypsin (15090-046, Thermo Fisher Scientific Inc.) and 0.02% (w/v) EDTA, and 10% of the cells from each well were transferred into a 10 cm dish. The next day, 1 mg/mL G418 Sulfate Solution (077-06433, FUJIFILM Wako Pure Chemical Corporation) was added to the cells, which were cultured until colonies formed approximately 10 d later. Colonies were picked up using a micropipette, transferred into 24-well plates, and cultured in medium with 1 mg/mL G418 until approximately 80% confluency was reached, then cells were dividedly transferred into two 12-well plates at different cell concentrations (10 and 90%). The former was used for Western blot analysis of TARGET-tagged VAT1 protein expression levels, and the latter was continuously cultured during Western blot analysis. Some clones highly expressing VAT1 were selected for further use.

Assay of NEP Activity

Sample preparation and the assay of NEP activity were carried out as described previously.¹¹⁾ Briefly, cells treated with 10 µM EGCg or 10 µM NUP-15 for 48 h were harvested and lysed in 40 µL lysis buffer A. Cell lysates were incubated in 100 mM MES buffer (pH 6.5) containing 0.1 mM succinyl-L-alanyl-L-alanyl-L-phenylalanine(Phe)-4-methyl-coumaryl-7-amide substrate (succinyl-Ala-Ala-Phe-MCA; Bachem, Bubendorf, Switzerland) at 37 °C. After 1 h incubation, 0.02 U/mL leucine aminopeptidase (Merck KGaA) and 10 µM phosphoramidon (Peptide Institute Inc.) were added to remove the Phe residue from Phe-MCA. The fluorescence intensity of released 7-amino-4-methylcoumarin was measured using excitation at 390 nm and emission at 460 nm. The NEP-dependent neutral endopeptidase activity was determined, based on the decrease in rate of digestion caused by 10 µM thiorphan (Merck KGaA), a specific inhibitor of NEP.

Statistical Analysis

Data are expressed as mean ± standard deviation. All statistical analyses were conducted using SigmaPlot software (ver. 14.0; Systat Software Inc., San Jose, CA, U.S.A.). Comparisons of mean values among more than three groups were conducted using two-way ANOVA and a post hoc Turkey’s test if the data passed the Shapiro–Wilk normality test and Brown–Forsythe equal variance test. Values of p < 0.05 were considered significant.

RESULTS AND DISCUSSION

Identification of VAT1 as a New EGCg-Binding Protein

To identify a new EGCg-binding protein, we synthesized EGCg derivatives with a C15 alkyl-chain spacer as bait compounds because we assumed that EGCg may not only bind with the protein surface but also be incorporated into or around a bioactive compound-binding pocket of any receptor. Addition of acrolein to the A-ring of EGCg enabled more facile direct binding of 15-hydroxypentadecanoic. Although this addition may affect the structure and recognition of EGCg, we believe that the B-, C- and G-rings of EGCg play more critical roles in its biological actions than the A-ring, as we reported previously.⁹⁾ In addition, if any non-specific protein binding occurred during the pull-down assay, these interactions would be excluded from subsequent SPR analysis of direct EGCg binding to the candidate protein. As a result, we obtained a pair of isomer derivatives with one C15 alkyl chain (NUP-E15-1a and -1b) and a derivative with two C15 alkyl chains (NUP-E15-2). We could not separate the isomers using chromatography, so we used NUP-E15-1 as a mixture for the following experiments.

We coupled NUP-E15-1, NUP-E15-2, and 15-hydroxypentadecanoic acid (negative control) to two kinds of magnetic beads via amine resides that were activated in advance using N-hydroxysulfosuccinimide and carbodiimide, then carried out a pull-down assay by mixing the membrane fraction from cell extracts. Sixteen protein bands containing candidate EGCg-binding proteins were pulled down by both NUP-E15-1 and -2 coupled to both types of magnetic beads, as determined using SDS-PAGE (Fig. 2, Supplementary Fig. 3). Using LC-MS/MS analysis, we identified 22 potential EGCg-binding proteins, the majority of which were mitochondria-related proteins that might have recognized the C15 alkyl chain of NUP-E15-1 and -2 as the alkyl chains of their ubiquinone and CoA substrates (Table 1). Likewise, retinol and prostaglandin H, which are the best substrates of retinol dehydrogenase and prostaglandin E synthase 2, respectively, have alkyl chains, and epoxide hydrolase 1 and sulfide:quinone oxidoreductase might have also recognized the catechol or pyrogallol structure of the B-ring of EGCg. Importantly, two previously reported molecules (eukaryotic translation initiation factor 3 subunit B and ATP synthase subunit beta, mitochondrial)¹²⁾ were identified as EGCg-binding candidates, although the identity of such EGCg-binding proteins may depend on the kind of cells or tissues used for screening. These results suggest that our pull-down assay worked. Finally, we selected VAT1 as an EGCg-binding protein candidate for further analysis for the following reasons: VAT1 is localized in synaptic vesicles and a part of presynaptic membranes,¹³⁾ suggesting that EGCg can to access and bind to VAT1 in cell extracts; the amino acid sequence of VAT1 shows partial homology with that of nicotinamide adenine dinucleotide phosphate reduced form (NADPH)-dependent quinone oxidoreductase,¹⁴⁾ suggesting that it may also recognize the catechol structure of EGCg; no cell surface receptor or related protein was detected in the pull-down assay, indicating that a protein involved in intracellular protein trafficking and protein stabilization may be a plausible candidate.

Fig. 2. Electrophoretic Patterns of Proteins Pulled Down Using EGCg-Coated Beads

Membrane fractions from human neuroglioma cells overexpressing human amyloid precursor protein with the Swedish mutation (H4-NL cells) were subjected to magnetic pull-down assays using (a) FG beads or (b) Dynabeads coated with either 15-hydroxypentadecanoic acid (ligand #1), NUP-E15-1 (ligand #2), or NUP-E15-2 (ligand #3). EGCg-binding protein candidates were eluted using buffers with different ionic strength and pH and separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) with 12.5 and 7.5% gels, respectively, and visualized using silver staining. The input samples (input) corresponded to the membrane fractions before pull-down. Green arrows indicate protein bands pulled down by both ligands #2 and #3 and both types of magnetic beads. The candidate proteins were extracted from the gels and analyzed using liquid chromatography–tandem mass spectrometry.

Table 1. Candidate Proteins Identified Using Mass Spectrometry

Band No.	Accession	Gene	Description	MW [kDa]	ΣCoverage	Σ# Unique peptides	Predicted binding site
1	Q16698	DECR1	2,4-Dienoyl-CoA reductase, mitochondrial	36	16.72	4	alkyl chain of ubiquinone
2	Q8TC12	RDH11	Retinol dehydrogenase	35.4	14.78	3	alkyl chain of retinol
	Q9H7Z7	PTGES2	Prostaglandin E synthase 2	41.9	10.61	4	alkyl chain of PGH
	Q16698	DECR1	2,4-Dienoyl-CoA reductase, mitochondrial	36	7.76	2	alkyl chain of ubiquinone
3	P05141	SLC25A5	ADP/ATP translocase 2	32.8	19.13	2	alkyl chain of CoA
	P12236	SLC25A6	ADP/ATP translocase 3	32.8	19.13	2	alkyl chain of CoA
	Q9H7Z7	PTGES2	Prostaglandin E synthase 2	41.9	5.31	2	alkyl chain PGH
4	P05141	SLC25A5	ADP/ATP translocase 2	32.8	13.76	2	alkyl chain of CoA
	P12236	SLC25A6	ADP/ATP translocase 3	32.8	13.76	2	alkyl chain of CoA
5*	P55884	EIF3B	Eukaryotic translation initiation factor 3 subunit B	92.4	6.88	5
6	P08865	RPSA	40S Ribosomal protein SA	32.8	7.12	2
8	O95202	LETM1	LETM1 and EF-hand domain-containing protein 1, mitochondrial	83.3	3.92	3	alkyl chain of CoA
9	P40939	HADHA	Trifunctional enzyme subunit alpha, mitochondrial	82.9	11.27	8	alkyl chain of ubiquinone
	O95573	ACSL3	Long-chain-fatty-acid-CoA ligase 3	80.4	4.72	3	alkyl chain of CoA
10	O00116	AGPS	Alkyldihydroxyacetonephosphate synthase, peroxisomal	72.9	12.61	7	alkyl chain
	P12956	XRCC6	X-ray repair cross-complementing protein 6	69.8	4.43	2
11	P49748	ACADVL	Very long-chain specific acyl-CoA dehydrogenase, mitochondrial	70.3	19.24	11	alkyl chain of ubiquinone
12&13	P61011	SRP54	Signal recognition particle 54 kDa protein	55.7	4.37	2
	P25705	ATP5F1A	ATP synthase subunit alpha, mitochondrial	59.7	4.16	2	alkyl chain of CoA
14*	P06576	ATP5F1B	ATP synthase subunit beta, mitochondrial	56.5	4.54	2	alkyl chain of CoA
15	Q99536	VAT1	Synaptic vesicle membrane protein VAT-1 homolog	41.9	25.45	6
	P07099	EPHX1	Epoxide hydrolase 1	52.9	14.29	7	catechol
	Q9Y6N5	SQRD	Sulfide:quinone oxidoreductase, mitochondrial	49.9	6.44	2	catechol
	P50454	SERPINH1	Serpin H1	46.4	5.74	2
	P39656	DDOST	Dolichyl-diphosphooligosaccharide--protein glycosyltransferase 48 kDa subunit	50.8	4.82	2
16	Q8NBX0	SCCPDH	Saccharopine dehydrogenase-like oxidoreductase	47.1	9.32	3
	P50454	SERPINH1	Serpin H1	46.4	5.74	2
	P07099	EPHX1	Epoxide hydrolase 1	52.9	4.18	2	catechol

No protein was detected in band #7. *Reported as an (−)-epigallocatechin-3-gallate (EGCg)-binding protein by Tanaka et al..¹²⁾

We prepared human VAT1 as a His-tagged recombinant protein using an E. coli protein expression system and purified it using fast protein liquid chromatography until a single protein band was obtained on a silver-stained SDS-PAGE gel (Supplementary Fig. 4). The kinetic interaction and affinity of EGCg or its derivatives with recombinant VAT1 were quantitatively analyzed using SPR. The sensorgrams showed a dose-dependent binding of NUP-E15-1, EGCg, and NUP-15 to the sensor chip-immobilized VAT1 (Fig. 3).

Fig. 3. Surface Plasmon Resonance Analysis of the Interaction between Recombinant Human Synaptic Vesicle Membrane Protein VAT-1 Homolog (VAT1) and Either EGCg, NUP-E15-1, or NUP-15

(a) Sensorgrams showing interaction between recombinant human VAT1 and either EGCg, NUP-E15-1, or NUP-15 in HBS-EP+ buffer (0.01 M HEPES, 0.15 M NaCl, 0.003 M EDTA, 0.05% (v/v) polysorbate 20, pH 7.4) containing 0.5% (v/v) DMSO. The interaction was analyzed at concentrations of 0, 1.25, 2.5, 5, 5, 10, and 20 µM (duplicate at 5 µM) for EGCg and its derivatives. The contact and dissociation times were set to 120 s and 300 s, respectively. The vertical axis quantifies in response units (RU) the amounts of compounds captured in each assay, and the horizontal axis shows the concentrations of compounds. *Although noise-like transient peaks at 0 s and 120 s were observed when buffer solutions with/without EGCg or its derivatives were introduced to the flow path, they did not affect the kinetic analysis. Values of kinetic parameters are summarized in Table 2. (b) Affinities of EGCg and its derivatives (1.25–20 µM) with His-tagged recombinant VAT1 protein immobilized via an anti-His-tag antibody on the sensor chip. Interactions were analyzed at a contact time of 120 s. The magnitude of the response at the time just before introducing the compound (i.e., the highest value) was plotted as a function of each compound concentration, and the K_D(affinity) values were calculated using Biacore T200 Evaluation Software. Average of K_D(affinity) values are summarized in Table 2. The equilibrium dissociation constant, K_D, was calculated from the ratio k_d/k_a. An index of the affinity between molecules, K_D(affinity), was calculated from the Michaelis–Menten equation.

Kinetic Study of Interaction Between VAT1 and EGCg

Interactions of NUP-E15-1, EGCg, and NUP-15 with VAT1 yielded k_a = (0.69–1.00) × 10³ M⁻¹s⁻¹, k_d = (6.8–9.4) × 10⁻⁴s⁻¹, and K_D = 0.98–1.20 µM; thus, no substantial difference among the compounds was observed. No sensorgram which passed the criteria of U-value in the interaction between VAT1 and other aliphatic catechins, NUP-6, -11, -18 or -19 in the present experimental condition, was obtained with good reproducibility. These results suggest that VAT1 recognizes a catechol or pyrogallol moiety around the B-, C-, and G-rings but not the long alkyl chains of EGCg (Fig. 3a, Table 2). Using K_m(affinity) as an index of binding affinity showed that the affinities of EGCg and NUP-15 binding to VAT1 were in the 10–20 µM range (Fig. 3b, Table 2). Although LC-MS/MS analysis and Western blotting after pull-down assays have revealed that a variety of proteins, receptors, and enzymes bind to EGCg,¹⁵⁾ it is often unclear whether they do so directly because one cannot exclude the possibility that other proteins may intermediate or assist the binding. To our knowledge, fewer than 15 binding proteins have been shown to bind directly to EGCg, as demonstrated using either a kinetic study with purified protein or binding assays involving radiolabeled EGCg, fluorescent protein, or bio-layer interferometry (Supplementary Table 1, refs contained therein).^4,5,16) VAT1 yielded an affinity for EGCg (K_D = 0.1–10 nM) that is comparable with the majority of EGCg-binding proteins, although vimentin and 67 kDa laminin receptor (67LR) have the most potent affinity for EGCg (K_D = 3–4 nM) (Supplementary Table 1, refs therein).^17,18) Compared with the rate constants for 67LR (k_a = 3.56 × 10⁴ M⁻¹s⁻¹, k_d = 1.42 × 10⁻³ s⁻¹), EGCg associates more slowly with VAT1 and dissociates faster.

Table 2. Comparison of Kinetic Parameters of Catechin Binding to Synaptic Vesicle Membrane Protein VAT-1 Homolog (VAT1)

Compounds	Association rate constant	Dissociation rate constant	Dissociation constant	KD(affinity) (µM)
	ka (M−1s−1)	kd (s−1)	KD (µM)
NUP-E15-1	1.00 × 103 ± 0.63	6.85 × 10−4 ± 4.64 × 10−4	0.98 ± 0.90	—*
EGCg	0.69 × 103 ± 0.02 × 103	7.39 × 10−4 ± 2.3 × 10−4	1.08 ± 0.36	19.50 ± 5.82
NUP-15	0.85 × 103 ± 0.27 × 103	9.32 × 10−4 ± 0.40 × 10−4	1.19 ± 0.40	13.33 ± 18.08

The constants quantifying the binding interaction of VAT1 with catechins were determined, and each experiment was repeated at least three times. Data represent the mean ± standard deviation. *Value was not determined because in two out of four experiments fitting curves of NUP-E15-1 did not reach the limiting rate at V_max in a Michaelis–Menten plot.

Biological Characterization of VAT1 Interactions with EGCg and NUP-15

We reported previously that some natural polyphenols, including EGCg, upregulate NEP activity and protein levels.⁹⁾ We therefore supposed that EGCg and NUP-15 effect this change by interacting with VAT1 in the cells. We prepared VAT1-overexpressing cells and examined the effects of these compounds on NEP upregulation. However, we did not further analyze using NUP-E15-1 and -2, because they showed a high cytotoxicity for the cultured cells. Treatment of mock cells with either EGCg or NUP-15 resulted in levels of both activity and protein content of NEP in the cell lysates that were up to approximately 40% greater than those in untreated mock cells, as observed previously (Iwata et al., unpublished data), whereas no effect was observed in the VAT1-overexpressing cells (Fig. 4). These unexpected results may indicate that VAT1 prevented the above effect by directly binding to and inactivating EGCg or NUP-15 in the cells overexpressing VAT1. Although detailed functions of VAT1 remain unclear, it will be interesting to explore whether cell migration in glioblastomas, where VAT1 is upregulated,^19,20) is affected by such a putative interaction.

Fig. 4. EGCg and NUP-15 Play a Role in NEP Retention to the Cell Surface by Interacting with VAT1

After H4-NL cells expressing human VAT1 and mock cells were treated with 10 µM EGCg or 10 µM NUP-15 for 48 h, their neprilysin (NEP) protein content (a, b) and activity (c) were analyzed. NEP band intensities in panel a were quantified using a LAS-4000 image analyzer and a standard curve constructed using diluted lysates. Data are shown as mean ± standard deviation (n = 4). Two-way ANOVA for changes in NEP content showed significant main effects of cell type (F_{(1, 18)} = 24.270; p < 0.001) and treatment (F_{(1, 18)} = 3.717; p = 0.045), with a significant interaction between cell type and treatment (F_{(1, 18)} = 7.305; p = 0.005). Two-way ANOVA for changes in NEP activity showed significant main effects of cell type (F_{(1, 18)} = 12.241; p = 0.003) and treatment (F_{(1, 18)} = 12.296; p < 0.001), with a significant interaction between cell type and treatment (F_{(1, 18)} = 3.852; p = 0.004). ^∗ p < 0.05.

CONCLUSION

We identified VAT1 as a new EGCg-binding protein with an affinity almost equivalent to that of previously reported binding proteins from neuroglioma cells. An interaction of VAT1 with EGCg and the aliphatic EGCg derivative NUP-15 may occur in the cells because the effects of these compounds were not observed when VAT1 was upregulated. Thus, we conclude that VAT1 is a new member of the EGCg-binding protein family.

Acknowledgments

This work was supported in part by the Japan Society for the Promotion of Science (JSPS) [Grant Number: 15K14971 to N.I.] and the Japan Agency for Medical Research and Development (AMED) [Grant Numbers: 16dk0207021h0001, 17dk0207021h0001, and JP18dk0207021h0003 to N.I.]. We thank Dr. Kei Maruyama for providing H4 cell lines (H4-NL and H4 BH-1). We also thank Eriko Koide for technical assistance.

Author Contributions

AI wrote the first draft of the paper; AI, NI, TT, KO, KS, and DH edited the paper; AI, NI, KS, and NK designed the research; AI, DH, KF, MH, KW, KO, TT, and KS performed the research; AI and NI analyzed the data and wrote the final version of the paper.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1) Khan N, Mukhtar H. Tea polyphenols in promotion of human health. Nutrients, 11, 39–54 (2018).
• 2) Chandrasekaran V, Hediyal TA, Anand N, Kendaganna PH, Gorantla VR, Mahalakshmi AM, Ghanekar RK, Yang J, Sakharkar MK, Chidambaram SB. Polyphenols, autophagy and neurodegenerative diseases: a review. Biomolecules, 13, 1196–1209 (2023).
• 3) Kochman J, Jakubczyk K, Antoniewicz J, Mruk H, Janda K. Health benefits and chemical composition of matcha green tea: a review. Molecules, 26, 85–95 (2020).
• 4) Negri A, Naponelli V, Rizzi F, Bettuzzi S. Molecular targets of epigallocatechin-gallate (EGCG): a special focus on signal transduction and cancer. Nutrients, 10, 1936–1959 (2018).
• 5) Tachibana H. Molecular basis for cancer chemoprevention by green tea polyphenol EGCG. Forum Nutr., 61, 156–169 (2009).
• 6) Zhong Y, Shahidi F. Lipophilized epigallocatechin gallate (EGCG) derivatives as novel antioxidants. J. Agric. Food Chem., 59, 6526–6533 (2011).
• 7) Wang M, Zhang X, Zhong YJ, Perera N, Shahidi F. Antiglycation activity of lipophilized epigallocatechin gallate (EGCG) derivatives. Food Chem., 190, 1022–1026 (2016).
• 8) Fudouji R, Tanaka T, Taguri T, Matsuo Y, Kouno I. Coupling reactions of catechins with natural aldehydes and allyl alcohols and radical scavenging activities of the triglyceride-soluble products. J. Agric. Food Chem., 57, 6417–6424 (2009).
• 9) Hori Y, Watanabe K, Yassen ASA, Shirotani K, Tanaka T, Iwata N. Enhancement of neprilysin activity by natural polyphenolic compounds and their derivatives in cultured neuroglioma cells. Biol. Pharm. Bull., 46, 446–454 (2023).
• 10) Seki T, Kanagawa M, Kobayashi K, Kowa H, Yahata N, Maruyama K, Iwata N, Inoue H, Toda T. Galectin 3–binding protein suppresses amyloid-β production by modulating β-cleavage of amyloid precursor protein. J. Biol. Chem., 295, 3678–3691 (2020).
• 11) Kawakubo T, Mori R, Shirotani K, Iwata N, Asai M. Neprilysin is suppressed by dual-specificity tyrosine-phosphorylation regulated kinase 1A (DYRK1A) in Down-syndrome-derived fibroblasts. Biol. Pharm. Bull., 40, 327–333 (2017).
• 12) Tanaka T, Ishii T, Mizuno D, Mori T, Yamaji R, Nakamura Y, Kumazawa S, Nakayama T, Akagawa M. (−)-Epigallocatechin-3-gallate suppresses growth of AZ521 human gastric cancer cells by targeting the DEAD-box RNA helicase p68. Free Radic. Biol. Med., 50, 1324–1335 (2011).
• 13) Linial M, Miller K, Scheller RH. VAT-1: an abundant membrane protein from Torpedo cholinergic synaptic vesicles. Neuron, 2, 1265–1273 (1989).
• 14) Kim SY, Mori T, Chek MF, Furuya S, Matsumoto K, Yajima T, Ogura T, Hakoshima T. Structural insights into vesicle amine transport-1 (VAT-1) as a member of the NADPH-dependent quinone oxidoreductase family. Sci. Rep., 11, 2120 (2021).
• 15) Moses MA, Henry EC, Ricke WA, Gasiewicz TA. The heat shock protein 90 inhibitor, (−)-epigallocatechin gallate, has anticancer activity in a novel human prostate cancer progression model. Cancer Prev. Res. (Phila), 8, 249–257 (2015).
• 16) Ermakova SP, Kang BS, Choi BY, Choi HS, Schuster TF, Ma WY, Bode AM, Dong Z. (−)-Epigallocatechin gallate overcomes resistance to etoposide-induced cell death by targeting the molecular chaperone glucose-regulated protein 78. Cancer Res., 66, 9260–9269 (2006).
• 17) Ermakova S, Choi BY, Choi HS, Kang BS, Bode AM, Dong Z. The intermediate filament protein vimentin is a new target for epigallocatechin gallate. J. Biol. Chem., 280, 16882–16890 (2005).
• 18) Tachibana H, Koga K, Fujimura Y, Yamada K. A receptor for green tea polyphenol EGCG. Nat. Struct. Mol. Biol., 11, 380–381 (2004).
• 19) Mertsch S, Becker M, Lichota A, Paulus W, Senner V. Vesicle amine transport protein-1 (VAT-1) is upregulated in glioblastomas and promotes migration. Neuropathol. Appl. Neurobiol., 35, 342–352 (2009).
• 20) Shan X, Wang K, Tong X, Wang Z, Wu F, Liu X, Yang P, Wang J. High expression of VAT1 is a prognostic biomarker and predicts malignancy in glioblastoma. Oncol. Rep., 42, 1422–1430 (2019).