Selective Intracellular Delivery of Ganglioside GM3 ... - ACS Publications

Jan 4, 2017 - Teruhiko Matsubara, Ryohei Otani, Miki Yamashita, Haruka Maeno, Hanae Nodono, and Toshinori Sato*. Department of Biosciences and ...
0 downloads 0 Views 7MB Size
Download PDF
Article pubs.acs.org/Biomac

Selective Intracellular Delivery of Ganglioside GM3-Binding Peptide through Caveolae/Raft-Mediated Endocytosis Teruhiko Matsubara, Ryohei Otani, Miki Yamashita, Haruka Maeno, Hanae Nodono, and Toshinori Sato* Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Kouhoku-ku, Yokohama 223-8522, Japan S Supporting Information *

ABSTRACT: Glycosphingolipids are major components of the membrane raft, and several kinds of viruses and bacterial toxins are known to bind to glycosphingolipids in the membrane raft. Since the viral genes and pathogenic proteins that are taken into cells are directly delivered to their target organelles, caveolae/raft-mediated endocytosis represents a promising pathway for specific delivery. In the present study, we demonstrated the ability of an artificial pentadecapeptide, which binds to ganglioside GM3, to deliver protein into cells by caveolae/raft-mediated endocytosis. The cellular uptake of a biotinylated GM3-binding peptide (GM3BP)−avidin complex into HeLa cells was observed, and the cellular uptake of this complex was inhibited by an incubation with sialic acid or endocytic inhibitors such as methyl-ß-cyclodextrin, and also by an incubation at 4 °C. These results indicate that the GM3BP-avidin complex bind to GM3 in membrane raft, and are taken into cell through caveolae/raft-mediated endocytosis. The GM3BP-avidin complex was transported into cells and localized around the nucleus more slowly than a human immunodeficiency virus type 1 TAT peptide. Furthermore, the uptake of a green fluorescent protein (GFP) linked with GM3BP into HeLa cells was similar to that of the GM3BP−avidin complex, and the localization of the GM3BP-GFP fusion protein was markedly different with that of the TAT-GFP fusion protein. The uptake and trafficking of GM3BP were distinguished from conventional cell-penetrating peptides. GM3BP has potential as a novel peptide for the selective delivery of therapeutic proteins and materials into cells in addition to being a cell-penetrating peptide.

1. INTRODUCTION Cell-penetrating peptides (CPPs), also referred to as protein transduction domains, are short peptides composed of basic amino acids and are able to transport pharmaceutical materials into mammalian cells.1 In 1988, transducing properties of human immunodeficiency virus type 1-encoded transactivator of transcription (TAT) protein were reported,2,3 and then it was indicated that a basic amino acid-rich peptide fragment of TAT protein, YGRKKRRQRRR, is responsible for the transduction of proteins into cells.4,5 The TAT peptide was fused with a biologically active protein ß-galactosidase (ß-Gal), and the TAT-ß-Gal fusion protein obtained was successfully internalized into cultured cells in vitro.5 Furthermore, in vivo experiments were performed, and ß-Gal activity was seen in all tissues of mice treated by an intraperitoneal injection.6 Together with the finding of translocation activity of TAT peptide, several CPPs have been reported such as Penetratin derived from Drosophila Antennapedia homeodomain (43−58 residues),7 herpes simplex virus VP22,8 murine vascular endothelial cadherin pVEC,9 and polyarginine.10 In the beginning of the study on TAT peptide, cellular uptake of TAT peptide into cells has been thought to be temperatureand energy-independent.4 The internalization of fluoresceinlabeled TAT peptide is observed at low temperature (4 °C) and in the presence of endocytosis inhibitors. This is because © XXXX American Chemical Society

C-terminal RRR residues interact with anionic proteoglycan and sialic acid on the cell surface11 and penetrate the cell membrane, therefore polyarginine (R9, etc.) is considered to act as CPPs.10 In addition, the GRKKR motif is known to act as a nuclear localization signal (NLS),12 and TAT peptide is able to transport cargoes to the nucleus.4 However, subsequently, energydependent endocytosis of TAT peptide has been reported. Dowdy et al. proposed that endocytosis of TAT-Cre fusion protein is involved in macropinocytosis.13 Furthermore, TATconjugated liposomes are internalized by energy-independent1 and energy-dependent pathways.11 The cellular uptake pathway may be influenced by the length and sequence of peptide, and the feature of cargoes.14,15 Membrane (lipid) raft is composed of sphingolipids and cholesterol and contributes to the formation and stability of caveolae. Caveolae are unique flask-shaped invaginations of the plasma membrane and are involved in cell signaling, lipid homeostasis, and membrane tension.16−18 Caveolar invaginations composed of caveolin proteins and membrane rafts are closed by dynamin, and caveolar vesicles are fused to early endosomes ranging in size from 50 to 80 nm. Cholera toxin B Received: August 23, 2016 Revised: December 12, 2016

A

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

fusion protein in DMEM/F12 at 4 or 37 °C for 30 min or longer. These cells were then washed with PBS, and treated with trypan blue (0.4%, w/v) to quench extracellular fluorescence.26,27 To observe the localization of the complex or fusion protein, cell fluorescence was analyzed using CLSM (TCS−NT, Leica) equipped with a Kr/Ar laser. 2.5. Flow Cytometry (FCM). The internalization of peptide−avidin complexes or the peptide-GFP fusion protein was determined by flow cytometric analysis. Cells (6 × 104 cells in 24-well plates) were grown at 37 °C for 24 h. After removal of the medium, cells were washed with PBS and incubated with the peptide complex with fluorescein-labeled avidin or peptide-GFP fusion protein in DMEM/F12 at 4 or 37 °C. These cells were then washed with PBS, harvested from the plates with 0.05% trypsin-EDTA, and washed with PBS. The cells were treated with trypan blue in PBS, and washed with 1% BSA/PBS. Cells were subjected to flow cytometric analysis (EPICS XL, Beckman Coulter). The fluorescence intensity of fluorescein-labeled avidin was subtracted from the intensity of the peptide−avidin complex. All experiments were carried out in triplicate. 2.6. Inhibition of Endocytosis by Inhibitors. HeLa cells were preincubated at 37 °C with 10 mg/mL methyl-ß-cyclodextrin (MßCD) for 60 min, 5 μg/mL filipin for 60 min, 10 μg/mL cytochalasin D for 30 min, 10 μM wortmannin for 60 min, or 10 μg/mL chlorpromazine for 30 min for inhibition experiments.28,29 Cells were then incubated with peptide−avidin complexes or the peptide-GFP fusion protein at 37 °C for 60 min, treated with trypan blue, and used for flow cytometric analysis as described above. 2.7. Intracellular Localization. HeLa cells seeded in glass-bottom dishes were preincubated with a 1:100 dilution of a FITC-conjugated anticaveolin-1 antibody or 10 μg/mL FITC-labeled transferrin at 37 °C for 30 min. After washing with DMEM/F12, cells were incubated with peptide−avidin complexes in DMEM/F12 at 37 °C for 30 min. Cells were then treated with trypan blue and used for the microscopic study, as described above. To stain lysosomes, HeLa cells were treated with LysoTracker Green or LysoTracker Red at 37 °C for 15 min after the incubation with peptide−avidin complexes or the peptide-GFP fusion protein for 30 min to 2 h. 2.8. Construction of Expression Vectors for Peptide-GFP Fusion Proteins. The vector pAcGFP1 (Clontech), which contained the GFP gene from Aequorea coerulescens, was developed for the construction of expression vectors. Using pAcGFP1 as a template, the GFP-h, carboxyl terminal (C-terminal) His6-tagged GFP, gene fragment was amplified using a primer pair 5′-CCA CCA TGG TGA GCA AGG-3′ and 5′-GTT GGA ATT CTA TCA TCA TCA CGC GGC CGC ATG ATG ATG ATG ATG ATG CTT GTA CAG CTC ATC CAT GCC-3′ (Figure S3a). The resultant PCR product was digested with NcoI/EcoRI and subcloned into pAcGFP1 to generate a plasmid named pAcGFP-h. The pAcGFP-h vector contained NotI and EcoRI restriction enzyme sites downstream of the AcGFP1 gene. To prepare the expression vectors of TAT peptide- and GM3BP-GFP fusion proteins, the oligonucleotide fragments encoding these peptides were synthesized, and insert duplexes were generated with the Klenow fragment using 5′-TGA GCG GCC GCG GGC GGC GGC GGC GGC TAT GGC CGT AAA AAA CGT CGT CAG CGT CGT CGT TGA TGA TGA GAA TTC CCG CCG-3′ for TAT or 5′-TGA GCG GCC GCG GGC GGC GGC GGC GGC GGC TGG TGG TAT AAA GGC CGT GCG CGT CCG GTG AGC GCG GTG GCG TGA TGA TGA GAA TTC CCG CCG-3′ for GM3BP with an extension primer 5′-CGGCGGGAATTCTCATCATCA-3′. The resultant duplex was digested and inserted into pAcGFP-h using NotI and EcoRI restriction sites to give the expression vectors pAcGFP-h-TAT and pAcGFP-hGM3BP (Figure S3b). These vectors in Escherichia coli could produce peptide-GFP fusion proteins, in which GFP was followed by a histidine hexamer (for purification), alanine trimer (EcoRI restriction site), glycine spacer, and target peptide (TAT peptide or GM3BP)(Table S1). The nucleotide sequences of GFP and peptide genes were confirmed in the expression vectors by DNA sequencing using the ABI PRISM Genetic Analyzer (Applied Biosystems). 2.9. Expression and Purification of Peptide-GFP Fusion Proteins. To express peptide-GFP fusion proteins, expression

subunit (CTB), which binds to ganglioside GM1 in the membrane raft, is therefore transported to endocytic organelles and colocalized with caveolin proteins.19 Simian virus 40 (SV40) is known to bind to GM1 and use the caveolae/raft-mediated pathway for infectious entry into cells with signal transduction including local protein tyrosine phosphorylation and depolymerization of the cortical actin cytoskeleton.20 Caveolae/raftmediated endocytosis may be one of the preferred pathways as well as other types of endocytosis for drug delivery systems.21 We previously selected pentadecapeptides that bind to ganglioside GM1,22,23 GM2,24 GM3,25 GD1a,24 and GT1b24 from a random peptide library using a phage display method. GM3-binding peptides (GM3BPs) identified from a selection against GM3 were previously shown to bind to MDCK cells and inhibit infection by the influenza virus.25 Cargoes conjugated to the ganglioside-binding peptide may be delivered into cells through the peptide binding. In the present study, a biotinylated GM3BP-avidin complex and a GM3BP-fused green fluorescent protein (GFP) were prepared to investigate the interaction between the GM3BP-linked protein and mammalian cells. GM3BP-linked proteins bound to sialylglycoconjugates including GM3 in the membrane raft, and were then taken into cells. GM3BP-linked proteins were selectively delivered into HeLa cells through caveolae/raft-mediated endocytosis; the route and distribution of GM3BP-linked proteins were distinguished from those of the TAT peptide. These results demonstrate that GM3BP has potential as a novel peptide that can deliver proteins through the caveolae/raft-mediated endocytic pathway.

2. MATERIALS AND METHODS 2.1. Materials. Restriction enzymes were purchased from New England Biolabs and TaKaRa (Tokyo, Japan). Peptide amides carrying a biotinyl group (biotinylated peptide) such as biotinyl-GM3BP (GWWYKGRARPVSAVAK(biotin)-NH2), biotinyl-TAT (YGRKKRRQRRRK(biotin)-NH2), and biotinyl-CP8 (AETVESCLAKPHTENK(biotin)-NH2) were chemically synthesized as described previously.25 Briefly, peptides were elongated with an automatic peptide synthesizer using standard 9-fluorenylmethyloxycarbonyl (Fmoc) chemistry, in which Fmoc-Lys(biotin)−OH was linked with the C-terminus of the peptide. The high purities (>98%) and expected structures of the biotinylated peptide amides were verified by reversed-phase highperformance liquid chromatography (HPLC) and matrix-assisted laser desorption ionization/time-of-flight mass spectrometry (MALDI-TOF MS), respectively (Supporting Information, Figure S1). The CP8 peptide was used as a control peptide, had a similar length of that of GM3BP, and was derived from the major coat protein VIII of the M13 phage. 2.2. Cell Culture. B16, COS7, and HeLa cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Biological Industries, Israel), 50−100 units/ml penicillin G, and 100 μg/mL streptomycin at 37 °C under 5% CO2−95% air. 2.3. Preparation of the Peptide−Avidin Complex. To prepare the peptide−avidin complex (molar ratio 4:1), a solution of biotinylated peptide (40 μM) was incubated with fluorescein (fluorescein isothiocyanate, FITC, or tetramethylrhodamine isothiocyanate, TRITC)labeled avidin (10 μM) in phosphate-buffered saline (PBS). The solution of the complexes was added to HeLa cells at a final concentration of 1 μM complex in DMEM/F12 (without phenol red) with or without inhibitors, where DMEM/F12 containing 1% bovine serum albumin (BSA) was available instead of DMEM/F12. 2.4. Confocal Laser Scanning Microscopy (CLSM). HeLa cells were seeded at 1−2 × 104 cells/dish in glass-bottom dishes (12 mm diameter) (IWAKI, Japan) and incubated for 24 h in DMEM containing 10% FBS. The cells were washed with PBS and incubated with the peptide complex with fluorescein-labeled avidin or the GFP-peptide B

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 1. Internalization of the GM3BP-avidin complex into HeLa cells. (a) Internalization of the GM3BP complex with FITC-avidin into HeLa cells. HeLa cells were incubated with 1 μM of the GM3BP−avidin complex at 37 °C for 10, 30, and 60 min. The bar represents 10 μm. (b) Inhibition of the internalization of the GM3BP−avidin complex into HeLa cells by sialic acid (Neu5Ac). HeLa cells were incubated with 1 μM of the peptide−avidin complex at 37 °C for 30 min in the absence (−) and presence of 200 mM Neu5Ac (+). CLSM image (left) and relative fluorescence intensities determined by flow cytometry (right). (c) Temperature dependence of the cellular uptake of the peptide−avidin complex. HeLa cells were incubated with 1 μM of the GM3BP complex at 4 or 37 °C for 30 min. (d) Influence of endocytosis inhibitors on cellular uptake. HeLa cells were preincubated with inhibitors at 37 °C for 30 or 60 min and incubated with the GM3BP−avidin complex at 37 °C for 60 min. Error bars indicate the standard deviation (n = 3). Statistical analysis was by a two-tailed unpaired Student’s t-test: *, p < 0.01; **, p < 0.001; ***, p < 0.0001. vectors were transformed into E. coli strain BL21(DE3) (BioDynamics Laboratory Inc.) (Tokyo, Japan). Cells were grown in Luria−Bertani (LB)/ampicillin (Amp) at 37 °C with shaking at 160 rpm to reach an OD600 of 0.5−0.7 prior to induction. To express GFP-h, TAT-GFP, and GM3BP-GFP, cells were stimulated with 0.5 mM isopropyl β-D-1thiogalactoside (IPTG) for several hours. The cells were then harvested, and fusion proteins were purified by metal chelate affinity chromatography using a HisTrap FF crude kit (GE Healthcare) according to the instruction manual. Briefly, a cell pellet was suspended in a binding buffer (20 mM phosphate buffer, 0.5 M NaCl, 20 mM imidazole, pH 7.4) and treated with DNase I, lysozyme, and protease inhibitors. After sonication, the soluble fraction was applied to the HisTrap FF crude column. The fusion protein was eluted with 500 mM imidazole, and imidazole in the eluate was removed by exchanging with PBS using a centrifugal filter unit (YM-10, Millipore). According to its absorbance at 280 nm, the solution of the fusion protein was typically 5−10 mg/mL. The purities of fusion proteins were confirmed by 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) stained with Coomassie blue (Figure S3c). The fluorescence of the fusion protein was measured at 510 nm (excitation at 475 nm) using a fluorescence photometer FL-2500 (Hitachi, Japan). To determine an apparent concentration (mol/L), the fluorescence intensities of the protein solution were interpolated onto the plot for the FITC standard curve. Typically, 1 mg/mL of the fusion protein solution estimated from its absorbance corresponded to an apparent concentration of 34 μM. The solution of fusion proteins was diluted to 1−20 μM, typically 5 μM (0.15 mg/mL), with DMEM/F12 (without phenol red) for CLSM and flow cytometric analyses. 2.10. Cytotoxicity of GM3BP against HeLa Cells. The sensitivity of HeLa cells to peptide was measured using the MTT assay with a cellcounting kit (Dojindo, Japan). HeLa cells (3 × 103/well) were cultured in 96-well microtiter plates in DMEM containing 10% FBS in the presence of the peptide−avidin complex. Cells were incubated at 37 °C for 24 h before the addition of a water-soluble tetrazolium salt (WST-1)/1-methoxy-5-methylphenazinium methylsulfate (1-methoxy PMS) solution for 2 h. Absorbance in each well was measured using a microplate reader at 450/690 nm. Cell viability was expressed as the ratio of absorbance at 450 nm of cells treated with the complex over control samples (without the complex).

3. RESULTS 3.1. Cellular Uptake of the GM3BP Complex with FITC-Labeled Avidin. The GM3 (Neu5Acα2−3Galβ1− 4Glcβ1−1′Cer)-binding pentadecapeptide (GM3BP, GWWYKGRARPVSAVAK), which was selected from a random peptide library, is known to bind to glycoconjugates containing sialylgalactose (sialylglycoconjugates) such as GM3, which is expressed on the surface of B16 and MDCK cells.25 In order to examine GM3BP-mediated cellular uptake, biotinylated peptides (Supporting Information, Figure S1) were combined with FITClabeled avidin, and the GM3BP−avidin complex was incubated with mouse B16, monkey COS7, and human HeLa cells. Internalization of the complex into the three cell lines was detected by flow cytometric studies (Figure S2a). HeLa cells were selected for further investigations because a large amount of internalization of the complex into cells was observed in the micromolar range. GM3 is a major component of gangliosides in HeLa cells.30 A CP8 peptide (AETVESCLAKPHTEN), which was used as the control peptide, bound to neither cell line.25 The fluorescein in CLSM images was identified as small dots near the cell surface after an incubation with the complex at 1 μM for 10 min and also inside cells within 30 min (Figure 1a). Binding of the GM3BP−avidin complex to cells was inhibited in the presence of sialic acid (Neu5Ac), which is consistent with previous findings (Figure 1b).25 These results demonstrated that the GM3BP complex bound to HeLa cells through the interaction between GM3BP and sialylglycoconjugates on the cell surface, followed by its rapid internalization into cells. In order to eliminate the influence of serum on the internalization, HeLa cells were incubated with the GM3BP complex in the presence of FBS. Significant decrease of fluorescein in CLSM images was not found after an incubation with the complex at 1 μM for 30 min in the presence of 10% FBS (Figure S2b). The temperature dependence of the binding and internalization of the peptide-avidin complex was examined using HeLa cells. Internalization of the GM3BP−avidin complex was markedly inhibited by incubation at 4 °C (Figure 1c). C

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules These results indicated that the GM3BP−avidin complex is internalized into cells by energy-dependent endocytosis. 3.2. Efficient Uptake of the GM3BP−Avidin Complex. To determine the efficiency of GM3BP uptake by HeLa cells, cells were incubated with the GM3BP complex (1 μM) and examined by flow cytometry. Fluorescence on the cell surface was quenched by treating cells with trypan blue to estimate the amount of the complex internalized into cells (Figure S2c). After incubation with HeLa cells at 37 °C for 60 min, more than 80% of the GM3BP complex bound to the cell surface was internalized into cells (FI+/FI− ratio, Table 1). Table 1. Relative Amount of GM3BP−Avidin Complexes (1 μM) and GM3BP-GFP Fusion Proteins (5 μM) Internalized into HeLa Cells FIa Trypan blue

GM3BP−avidin complex

GM3BP-GFP fusion protein

a

incubation time (min)

−

+

FI+/FI− ratio

10 30 60 90 10 30 60 90

25 57 93 127 2.1 4.3 11 17

19 46 75 113 2.1 3.8 7.5 15

0.76 0.81 0.81 0.89 1.0 0.88 0.68 0.88

Figure 2. Influence of endocytosis inhibitors on cellular uptake of the GM3BP-avidin complex. HeLa cells were preincubated with inhibitors at 37 °C for 30 or 60 min, and incubated with the GM3BP−avidin complex at 37 °C for 60 min.

In order to trace intracellular trafficking of the GM3BP−avidin complex, lysosomes were stained with LysoTracker Green. After the incubation of the GM3BP avidin complex at 37 °C for 30 min to 2 h, CLSM images showed the absence of yellow dots (Figure 3c). The GM3BP−avidin complex was not transported to lysosomes in HeLa cells within 2 h. After 4 h or more, the GM3BP avidin−complex (red dots) gradually accumulated around the outside of the nucleus (Figure 3d). The results obtained indicated that the GM3BP−avidin complex remained in endosomes after 24 h. 3.5. Preparation of Peptide-GFP Fusion Proteins. In order to clarify the interaction of GM3BP-linked cargoes with cells, a GM3BP-GFP fusion protein (GM3BP-GFP) in E. coli was prepared in the present study. Yang et al. used the TAT peptide-GFP fusion protein (TAT-GFP) to examine the intracellular transduction and cellular localization of TAT-GFP in mammalian cells.34 GM3BP and the TAT peptide were linked to the carboxyl terminus of GFP-h to give GM3BP-GFP and TAT-GFP, in which GFP-h is a histidine tag (His6) containing GFP to purify peptide-GFP fusion proteins with a Ni2+ affinity column (Figure 4a and Table S1). The expression vectors of the peptide-GFPs were constructed (Figures S3a and S3b), and then transformed into E. coli strain BL21 (DE3) cells to express fusion proteins as reported previously.34 The peptide-GFPs were expressed as soluble proteins, and then purified using metal chelate affinity chromatography. A single band was confirmed at approximately 30 kDa by SDS-PAGE with CBB staining (Figure S3c). The fluorescence intensity of the peptide-GFP solution was measured and compared to the fluorescence intensity of a series of diluted FITC solutions to determine the apparent concentration (mol/L). 3.6. Cellular Uptake of the GM3BP-GFP Fusion Protein. When the GM3BP-GFP fusion protein (5 μM) was incubated with HeLa cells at 37 °C for 60 min, the internalization of GM3BP-GFP that bound to the cell surface was observed in addition to that of the GM3BP−avidin complex (Figure 4b, Table 1). The fluorescence intensity of an unmodified control GFP (GFP-h) bound to the cell surface was weaker than that of GM3BP-GFP; therefore, GM3BP-mediated internalization was indicated. To determine whether the internalization of GM3BP-GFP into cells was energy-dependent, HeLa cells were incubated with GM3BP-GFP at 4 and 37 °C. The cellular uptake of

Fluorescence intensity.

3.3. Cellular Uptake of the GM3BP−Avidin Complex by Caveolae/Raft-Mediated Endocytosis. HeLa cells were pretreated with endocytosis inhibitors to determine the cellular uptake pathway of the GM3BP complex. Since GM3BP is able to bind to sialylglycoconjugates such as ganglioside GM3 and glycoproteins, the membrane microdomain such as membrane rafts is considered to contribute to cellular uptake. Pretreatments with MβCD and filipin, which inhibit caveolae/raft-mediated endocytosis,31 blocked the cellular uptake of the GM3BP−avidin complex by 85% and 45%, respectively (Figures 1d and 2), whereas cytochalasin D and wortmannin, which are inhibitors of macropinocytosis,32,33 only slightly blocked its uptake (22% and 12%, respectively). Chlorpromazine, which is a clathrin-mediated endocytosis inhibitor,32 showed no significant inhibition. The impact of the inhibitors on cell viability was negligible for HeLa cells in the present condition (Figure S2d). These results suggest that caveolae/raft-mediated endocytosis is the main pathway for the cellular uptake of the GM3BP complex. 3.4. Subcellular Localization of the GM3BP−Avidin Complex. Co-localization of the TRITC-labeled complex with a FITC-conjugated anticaveolin-1 antibody was observed by CLSM to determine intracellular trafficking of the GM3BP avidin complex. Figure 3a shows that yellow dots were observed in cells after 30 min. These results indicated that the GM3BP− avidin complex colocalized with caveolin-1. On the other hand, the GM3BP avidin complex did not colocalize with FITC-labeled transferrin as a marker for clathrin-mediated endocytosis (Figure 3b). These results were consistent with those obtained for the effects of endocytosis inhibitors shown in Figure 1d. Therefore, the GM3BP avidin complex is considered to be mainly internalized into cells by caveolae/raft-mediated endocytosis. D

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 3. Intracellular localization of the GM3BP−avidin complex and TAT-avidin complex in HeLa cells. (a) Co-localization of the GM3BP−avidin complex with caveolin-1. HeLa cells were incubated with 1 μM of the GM3BP complex with TRITC-avidin at 37 °C for 30 min after an incubation with a FITC-labeled anticaveolin-1 antibody. (b) Subcellular distribution of the GM3BP−avidin complex and transferrin. HeLa cells were incubated with the GM3BP complex at 37 °C for 30 min after the incubation with FITC-transferrin (overlay). (c) Subcellular distribution of the GM3BP−avidin complex and lysosomes. HeLa cells were incubated with the GM3BP complex at 37 °C for 30 min, 1 h, and 2 h before the incubation with LysoTracker Green (overlay). (d) Intracellular localization of the GM3BP−avidin complex after a 4−24 h incubation. HeLa cells were incubated with the GM3BP complex at 37 °C for 4 and 24 h. (e) Co-localization of the TAT peptide−avidin complex with lysosomes in HeLa cells. HeLa cells were incubated with 1 μM of the TAT complex with TRITC-avidin at 37 °C for 1 and 2 h before the incubation with LysoTracker Green (overlay). White arrows indicate the colocalization of the TAT complex with lysosomes. [GM3BP complex or TAT complex] = 1 μM. The bar represents 20 μm.

Figure 4. Internalization of the GM3BP-GFP fusion protein into HeLa cells. (a) Construction of the GM3BP-GFP fusion protein (GM3BP-GFP) and a histidine-tagged control GFP (GFP-h). GM3BP was linked to the carboxyl terminal region of GFP-h. (b) Internalization of GM3BP-GFP with HeLa cells. HeLa cells were incubated with 5 μM GM3BP-GFP or GFP-h at 37 °C for 60 min ([peptide-GFP fusion protein] = 5 μM, 0.15 mg/mL). The bar represents 10 μm. (c) Temperature dependence of the cellular uptake of GM3BP-GFP. HeLa cells were incubated with 5 μM GM3BP-GFP at 4 or 37 °C for 60 min, and subjected to a flow cytometric analysis. (d) Influence of endocytosis inhibitors on cellular uptake. HeLa cells were preincubated with inhibitors at 37 °C for 30 or 60 min and incubated with GM3BP-GFP for 60 min. Error bars indicate the standard deviation (n = 3). Statistical analysis was by a two-tailed unpaired Student’s t-test: *, p < 0.01; **, p < 0.001; ***, p < 0.0001.

GM3BP-GFP was significantly reduced at 4 °C (Figure 4c). In addition, the uptake of GM3BP-GFP was inhibited after cells were pretreated with MβCD; its uptake efficiency decreased to 31% (Figures 4d and S4). The MβCD-dependent internalization of GM3BP-GFP was largely inhibited as was observed with the GM3BP complex (Figure 1d). Cytochalasin D and chlorpromazine

blocked 34% and 28% of the cellular uptake of GM3BP-GFP, respectively. These results indicated that GM3BP-GFP was also uptaken by caveolae/raft-mediated endocytosis as was observed with the GM3BP complex. In order to determine the subcellular distribution of GM3BPGFP, lysosomes were stained using LysoTracker Red after the E

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 5. Intracellular localization of GM3BP-GFP (upper) and TAT-GFP (lower) in HeLa cells. HeLa cells were incubated with 1 μM of the GM3BP-GFP or TAT-GFP at 37 °C for 60 min after an incubation with LysoTracker Red. The bar represents 10 μm.

incubation of GM3BP-GFP at 37 °C for 60 min. CLSM images indicated that GM3BP-GFP did not localize in lysosomes, whereas TAT-GFP was largely transported to lysosomes (Figure 5). 3.7. Cellular Uptake of the TAT Peptide-Linked Protein by Macropinocytosis. In order to investigate the cellular uptake pathway of the TAT peptide-linked protein, HeLa cells were pretreated with endocytosis inhibitors. Regarding the TAT peptide−avidin complex, a pretreatment with MβCD, filipin, cytochalasin D, and wortmannin resulted in the significant inhibition of cellular uptake in the same experiments (44%, 22%, 62%, and 36% inhibition)(Figure S5a). The TAT complex was taken up by a combination of macropinocytosis (major) and caveolae/raft-mediated endocytosis (minor) as reported previously.35 CLSM images indicated that the TAT complex was localized in lysosomes after 1−2 h, which supports the uptake of the complex by macropinocytosis (Figure 3e). Similar results were obtained with the inhibition of the cellular uptake of TAT-GFP using treatments with endocytosis inhibitors (Figure S5b); TAT-GFP was also mainly taken up by macropinocytosis similar to the TAT complex. 3.8. Peptide Cytotoxicity. The cytotoxicity of GM3BP was investigated after an incubation of HeLa cells with the peptide− avidin complex at concentrations up to 10 μM for 24 h. A decrease in cell viability was not observed following a 24-h incubation, even at 10 μM, with GM3BP and the TAT complex (Figure S2e) (Student’s two-tailed unpaired t-test, P > 0.05, n = 4). Therefore, we concluded that the peptide−avidin complex does not exhibit toxicity.

selectivity for target cells or endocytic pathways. Many types of tumor-associated gangliosides have been detected in the plasma membrane, and these gangliosides are attractive targets for cancer immunotherapies.38 Short peptides that bind to cell surface receptors, at least in principle, have the characteristic features of CPPs. Therefore, we considered ganglioside-binding peptides to function as CPPs with cell selectivity. We previously reported the binding of GM3BP, GWWYKGRARPVSAVA, to sialylglycoconjugates on B16 and MDCK cell surfaces.25 When the biotinylated GM3BP−avidin complex was incubated with three ganglisoside-containing cells such as mouse B16, monkey COS7, and human HeLa cells, internalization of the GM3BP complex was clearly observed (Figures 1a and S2a). Since GM3 is a major component of gangliosides in HeLa cells,30 HeLa cells were used in further investigations. Although the amount of internalization of the GM3BP complex into HeLa cells was less than that of the TAT complex, sufficient amounts of the complex were internalized into HeLa cells (Figures 1a and S2a). Internalization of the GM3BP complex was inibited by the incubation at 4 °C, which indicated energy-dependent endocytosis of the complex (Figure 1c). The cellular uptake of the GM3BP avidin complex was determined using FCM and CLSM. The pretreatment of cells with raft/caveolae inhibitors (MßCD and filipin)(Figures 1d and 2) and double staining with the anticaveolin-1 antibody (Figure 3a) indicated that caveolae/raftmediated endocytosis was the main pathway for internalization of the GM3BP avidin complex (Figure 6). This internalization was inhibited in the presence of sialic acid (Figure 1b), which supports the internalization of GM3BP through endocytosis being mediated by the binding of GM3BP to sialylglycoconjugates on the cell surface. On the other hand, internalization of the TAT complex was markedly reduced in the presence of inhibitors of macropinocytosis (cytochalasin D and wortmannin), indicating that cellular uptake of the TAT complex mainly occurred via macropinocytosis (Figure S5). The different routes of endocytosis between GM3BP and the TAT peptide are considered to be due to differences in the target molecules on the cell surface. GM3BP-GFP showed similar uptake and trafficking to the GM3BP complex: the strong inhibition of internalization in the

4. DISCUSSION CPPs are short peptides composed of basic amino acid-rich peptide fragments that transport pharmaceutical materials into mammalian cells.1 The TAT peptide interacts with anionic proteoglycans, e.g., heparin and dextran sulfate,11 and TATlinked cargoes are taken into cells by macropinocytosis and transported to lysosomes.13,15 Although CPPs are powerful tools to deliver large amounts of cargoes to cells, the release of cargoes from lysosomes transported by the TAT peptide is required for their delivery to the cytoplasm/nucleus without degradation.36,37 In addition, conventional CPPs such as the TAT peptide have no F

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 6. Schematic diagram of the selective cellular uptake of the GM3BP-linked protein into HeLa cells. The GM3BP-linked protein binds with sialylglycoconjugates on the cell surface and is internalized into cells by caveolae/raft-mediated endocytosis. The uptake and trafficking of GM3BP were distinguished from those of the TAT peptide; GM3BP-linked proteins require more time to be transported to lysosomes than TAT-linked proteins.

■

presence of raft/caveolae inhibitors (Figure 4) and localization in endosomes (Figure 5). A marked difference was observed in localization between GM3BP-GFP and TAT-GFP after a 60 min incubation. This result indicates that when cargoes are linked to GM3BP, the cellular uptake and trafficking of GM3BP are clearly distinguishable from those of the TAT peptide. GM3BP has potential as a CPP mediated by caveolae/raft-mediated endocytosis with a different route to that of conventional CPPs. The escape of CPP conjugates from endosomes to the cytoplasm and/or nucleus is required, e.g., endosome-disrupting agents and a Cre-recombinase system have been used.37 Since GM3BP-linked proteins require more time to be transported to lysosomes than TAT-linked proteins (Figures 3, 5, and 6), the continuous endosomal release of cargoes may be expected when conjugated with endosome-disrupting agents. After the endosomal release of cargoes, nuclear entry during mitosis is also expected. Caveolar trafficking may act as an effective delivery system using the same mechanism as the CPP system.39 Oh et al. reported that aminopeptidase P (APP) was abundant in the caveolae of the lung endothelium, and an anti-APP antibody induced effective transendothelial transport in vivo.40 Some pathogens, such as SV and CTB, exploit caveolae/raft-mediated endocytosis to escape lysosomal degradation.41 GM3BP can easily be linked to various cargoes such as drugs, vectors, proteins, and fluorescent agents. Together with caveolar ligands, membrane raft-targeting molecules are promising carriers for efficient drug delivery and gene transfection.42

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.6b01262. Typical HPLC chromatogram and mass spectrum of biotinylated peptide amides (Figure S1), interaction of the GM3BP-avidin complex with cells (Figure S2), list of peptide-GFP fusion proteins (Table S1), construction of the expression vectors and SDS-PAGE analysis of peptideGFP fusion protein (Figure S3), influence of endocytosis inhibitors on the cellular uptake of the GM3BP-GFP fusion protein (Figure S4), and TAT-linked proteins (Figure S5) (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Address: Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Kouhoku-ku, Yokohama 223-8522, Japan; Tel.:+81-45-566-1771; E-mail: [email protected]. ORCID

Teruhiko Matsubara: 0000-0002-8006-4324 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This study was partly supported by KAKENHI Grant Numbers JP24650283 (T.S.) and JP17750166 (T.M.), and a grant for the Keio Gijuku Fukuzawa Memorial Fund for the Advancement of Education and Research (T.M.).

■

CONCLUSION We investigated cellular uptake mediated by the gangliosidebinding peptide GM3BP. The GM3BP−avidin complex and GM3BP-GFP fusion protein bound to sialylglycoconjugates on the cell surface and were subsequently taken into cells through caveolae/raft-mediated endocytosis. The uptake and trafficking of GM3BP were distinguished from those of cell-penetrating peptides; the GM3BP-linked proteins required more time to be transported to lysosomes than TAT-linked proteins, and, thus, the continuous endosomal release of cargoes may be expected. Our results indicate that GM3BP has potential as a novel peptide that delivers pharmaceutical cargoes into cells through caveolae/ raft-mediated endocytosis.

■

REFERENCES

(1) Torchilin, V. P.; Rammohan, R.; Weissig, V.; Levchenko, T. S. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 8786−8791. (2) Green, M.; Loewenstein, P. M. Cell 1988, 55, 1179−1188. (3) Frankel, A. D.; Pabo, C. O. Cell 1988, 55, 1189−1193. (4) Vives, E.; Brodin, P.; Lebleu, B. J. Biol. Chem. 1997, 272, 16010− 16017. (5) Nagahara, H.; Vocero-Akbani, A. M.; Snyder, E. L.; Ho, A.; Latham, D. G.; Lissy, N. A.; Becker-Hapak, M.; Ezhevsky, S. A.; Dowdy, S. F. Nat. Med. 1998, 4, 1449−1452. G

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules (6) Schwarze, S. R.; Ho, A.; Vocero-Akbani, A.; Dowdy, S. F. Science 1999, 285, 1569−1572. (7) Gratton, J. P.; Yu, J.; Griffith, J. W.; Babbitt, R. W.; Scotland, R. S.; Hickey, R.; Giordano, F. J.; Sessa, W. C. Nat. Med. 2003, 9, 357−362. (8) Bennett, R. P.; Dalby, B.; Guy, P. M. Nat. Biotechnol. 2002, 20, 20. (9) Elliott, G.; O’Hare, P. Cell 1997, 88, 223−233. (10) Futaki, S.; Suzuki, T.; Ohashi, W.; Yagami, T.; Tanaka, S.; Ueda, K.; Sugiura, Y. J. Biol. Chem. 2001, 276, 5836−5840. (11) Console, S.; Marty, C.; Garcia-Echeverria, C.; Schwendener, R.; Ballmer-Hofer, K. J. Biol. Chem. 2003, 278, 35109−35114. (12) Ruben, S.; Perkins, A.; Purcell, R.; Joung, K.; Sia, R.; Burghoff, R.; Haseltine, W. A.; Rosen, C. A. J. Virol. 1989, 63, 1−8. (13) Wadia, J. S.; Dowdy, S. F. Adv. Drug Delivery Rev. 2005, 57, 579− 596. (14) Gump, J. M.; Dowdy, S. F. Trends Mol. Med. 2007, 13, 443−448. (15) Zorko, M.; Langel, U. Adv. Drug Delivery Rev. 2005, 57, 529−545. (16) Echarri, A.; Del Pozo, M. A. J. Cell Sci. 2015, 128, 2747−2758. (17) Hayer, A.; Stoeber, M.; Ritz, D.; Engel, S.; Meyer, H. H.; Helenius, A. J. Cell Biol. 2010, 191, 615−629. (18) Nabi, I. R.; Le, P. U. J. Cell Biol. 2003, 161, 673−677. (19) Parton, R. G. J. Histochem. Cytochem. 1994, 42, 155−166. (20) Pelkmans, L.; Helenius, A. Traffic 2002, 3, 311−320. (21) Brock, R. Bioconjugate Chem. 2014, 25, 863−868. (22) Matsubara, T.; Ishikawa, D.; Taki, T.; Okahata, Y.; Sato, T. FEBS Lett. 1999, 456, 253−256. (23) Matsubara, T.; Iida, M.; Tsumuraya, T.; Fujii, I.; Sato, T. Biochemistry 2008, 47, 6745−6751. (24) Matsubara, T.; Onishi, A.; Sato, T. Bioorg. Med. Chem. 2012, 20, 6452−6458. (25) Matsubara, T.; Sumi, M.; Kubota, H.; Taki, T.; Okahata, Y.; Sato, T. J. Med. Chem. 2009, 52, 4247−4256. (26) Rejman, J.; Bragonzi, A.; Conese, M. Mol. Ther. 2005, 12, 468− 474. (27) Loike, J. D.; Silverstein, S. C. J. Immunol. Methods 1983, 57, 373− 379. (28) Walther, C.; Meyer, K.; Rennert, R.; Neundorf, I. Bioconjugate Chem. 2008, 19, 2346−2356. (29) Hagiwara, K.; Nakata, M.; Koyama, Y.; Sato, T. Biomaterials 2012, 33, 7251−7260. (30) Markwell, M. A.; Fredman, P.; Svennerholm, L. Biochim. Biophys. Acta, Biomembr. 1984, 775, 7−16. (31) Parton, R. G.; Richards, A. A. Traffic 2003, 4, 724−738. (32) Perumal, O. P.; Inapagolla, R.; Kannan, S.; Kannan, R. M. Biomaterials 2008, 29, 3469−3476. (33) Araki, N.; Johnson, M. T.; Swanson, J. A. J. Cell Biol. 1996, 135, 1249−1260. (34) Yang, Y.; Ma, J.; Song, Z.; Wu, M. FEBS Lett. 2002, 532, 36−44. (35) Wadia, J. S.; Schaller, M.; Williamson, R. A.; Dowdy, S. F. PLoS One 2008, 3, e3314. (36) Huang, Y.; Jiang, Y.; Wang, H.; Wang, J.; Shin, M. C.; Byun, Y.; He, H.; Liang, Y.; Yang, V. C. Adv. Drug Delivery Rev. 2013, 65, 1299− 1315. (37) Wadia, J. S.; Stan, R. V.; Dowdy, S. F. Nat. Med. 2004, 10, 310− 315. (38) Krengel, U.; Bousquet, P. A. Front. Immunol. 2014, 5, 325. (39) Copolovici, D. M.; Langel, K.; Eriste, E.; Langel, U. ACS Nano 2014, 8, 1972−1994. (40) Oh, P.; Borgstrom, P.; Witkiewicz, H.; Li, Y.; Borgstrom, B. J.; Chrastina, A.; Iwata, K.; Zinn, K. R.; Baldwin, R.; Testa, J. E.; Schnitzer, J. E. Nat. Biotechnol. 2007, 25, 327−337. (41) Carver, L. A.; Schnitzer, J. E. Nat. Rev. Cancer 2003, 3, 571−581. (42) Sahay, G.; Alakhova, D. Y.; Kabanov, A. V. J. Controlled Release 2010, 145, 182−195.

H

DOI: 10.1021/acs.biomac.6b01262 Biomacromolecules XXXX, XXX, XXX−XXX