Efficient Mini-Transporter for Cytosolic Protein ... - ACS Publications

Sep 15, 2016 - ABSTRACT: An efficient method to deliver active proteins into cytosol is highly desirable to advance protein-based therapeutics. Argini...
0 downloads 0 Views 832KB Size
Subscriber access provided by UNIVERSITY OF LEEDS

Article

An Efficient Mini-Transporter for Cytosolic Protein Delivery Ning Zhang, Ziqiang Yan, Xue Zhao, Qing Chen, and Mingming Ma ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b08202 • Publication Date (Web): 15 Sep 2016 Downloaded from http://pubs.acs.org on September 16, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

An Efficient Mini-Transporter for Cytosolic Protein Delivery Ning Zhang, Ziqiang Yan, Xue Zhao, Qing Chen, and Mingming Ma* CAS Key Laboratory of Soft Matter Chemistry, iChEM (Collaborative Innovation Center of Chemistry for Energy Materials), Department of Chemistry, University of Science and Technology of China, Hefei, Anhui 230026, P.R. China ABSTRACT: An efficient method to deliver active proteins into cytosol is highly desirable to advance protein-based therapeutics. Arginine-rich cell-penetrating peptides (RPPs) have been intensively studied for intracellular protein delivery, and their applications require further improvement on delivery efficiency, serum stability and cytotoxicity. Designing synthetic analogs of RPPs provides an alternative way to achieve efficient cytosolic protein delivery. Herein we report the design and synthesis of a dendritic small molecule TG6, which is composed of one rigid planar core and four flexible arms with one guanidinium on each arm. Protein structure and function are well preserved in the TG6-protein conjugates, which are readily internalized into cytosol. Our study demonstrates that TG6 is a serum-stable and low-toxic molecular transporter delivering both small cargoes and large active proteins efficiently into cytosol. KEYWORDS: Protein delivery • Guanidinium • Cell penetrating peptides • Cytosolic delivery • Structure design

INTRODUCTION Functional proteins, including cytokines, antibodies and enzymes, have been widely investigated for the development of protein-based therapeutics.1 Although many proteins that target intracellular biological activity are considered as promising therapeutics, the delivery of active proteins efficiently into cytosol remains a big challenge, due to the large sizes, charged surfaces and inherent instability of active proteins.1 Various transporters such as cationic peptides,2-5 engineered proteins,6 synthetic polymers,7-9 lipidoids,10-12 boronates,13 oligonucleotides,14 nanocapsules,15 nanogels,16-17 carbon-based nano18-19 materials and inorganic nanoparticles20-21 have been developed for intracellular protein delivery. Among these transporters, arginine-rich cell-penetrating peptides (RPPs), are capable of crossing the cell membrane and transporting cargoes at low micromolar concentrations without causing significant membrane damage (termed transduction),22 which offers the cargoes immediate access to cytosol.23 The transduction efficiency of RPPs is affected by the size of cargoes, which has confined the application of RPPs to the transduction of small cargoes such as molecular probes and peptides.24 Large cargoes such as proteins coupled to linear RPPs are mainly taken up into endocytic vesicles, leading to endosomal trapping and lysosomal degradation.24 Recent studies have demonstrated enhanced transduction of cyclic RPPs25-27 and stapled RPPs28 over corresponding linear RPPs. In addition, cyclic TAT has been utilized as a covalent transporter for delivering small proteins such as GFP into cytosol.29 Besides the delivery efficiency, the application of RPPs requires further improvements on their serum stability30 and cytotoxicity.31-32 To enhance the serum stability and explore the structure diversity, various syn-

thetic analogs bearing guanidinium groups (Gdms) have been developed.33-37 Similar to RPPs, at least 6 Gdms are required in these analogs to show transduction activity,3338 which could lead to considerable cytotoxicity due to the high positive charge from multiple Gdms.33 Designing novel synthetic analogs of RPPs provides an alternative way to achieve cytosolic protein delivery, but it remains a challenge to reach the three goals (high efficiency for cytosolic protein delivery, high serum stability and low cytotoxicity) simultaneously with synthetic transporters.33, 36 To address this challenge, we hypothesize that a RPPmimicking non-peptide molecule with a rigid core and static presentation of a low number of Gdms around the rigid core could achieve the three goals simultaneously. Therefore, we have rationally designed dendritic small molecules containing only 4 Gdms, which are called Tetra-Guanidinium molecules (TGn, n = 3 or 6, indicating the number of methylenes in each arm, Scheme 1, S1-S2). TGn was covalently conjugated to small cargoes (fluorescein) or proteins (bovine serum albumin and horseradish peroxidase), and evaluated as transporters for cytosolic delivery. One of them, TG6, has been found as a serumstable and low-toxic transporter for efficient delivery of both small cargoes and large active proteins into cytosol.

EXPERIMENTAL SECTION Detailed synthesis and characterization of all the small molecules are described in the SI. Synthesis of BSA-FITC. BSA-FITC was prepared by mixing 1 equivalent of bovine serum albumin (BSA) with 1.5 equivalent of fluorescein isothiocyanate (FITC) in 100 mM NaHCO3 buffer (pH 8.5). The solution was gently mixed at room temperature for 1 h, and then purified by three times of ultrafiltration using Amicon Ultra-15 cen-

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

trifugal filter units with 10 kDa cutoff (Millipore) against PBS buffer. Synthesis of TG6-BSA-FITC and TG6-HRP. 2 mg TG6, 2 mg 2-(1H-benzotriazol-1-yl)-1,1,3,3,-tetramethyluronium hexa-fluorophosphate (HBTU) and 3 mg Nhydroxysuccimide (NHS) were dissolved in 0.1 mL N,Ndimethyl-formamide (DMF) and stirred at room temperature for 1 h. BSA-FITC or horseradish peroxidase (HRP) was dissolved in 4 mL PBS buffer. The activated TG6 in 0.1 mL DMF was first diluted with 0.4 mL PBS buffer, and then added to the BSA-FITC or HRP solution at a molar ratio of TG6:protein = 5:1. The solution was gently mixed at room temperature for 2 h, and then purified by three times of ultrafiltration using Amicon Ultra-15 centrifugal filter units with 10 kDa cutoff (Millipore) against PBS buffer. Characterization of modified proteins. Modified proteins were analyzed by a MALDI-TOF (Bruker) with sinapic acid as the matrix. Circular Dichroism (CD) spectra of proteins were recorded on a Jasco J-810 spectropolarimeter. Proteins were diluted to 0.5 mg/mL in PBS and analyzed in a 0.1-cm path length quartz cell (for 200-250 nm region) or in a 1-cm path length quartz cell (for 250450 nm region). All CD spectra were baseline-corrected with PBS as the blank, and averaged based on three replicates. Both of hydrodynamic diameter and zeta potential of proteins were determined by using a Zetasizer NanoZS900 (Malvern). Proteins were diluted to 5 mg/mL in PBS buffer (pH 7.4), filtered through a 0.2 μm filter, and then analyzed. All the results were averaged based on three replicates. Enzyme activity of HRP and TG6-HRP. 0.5 μg/mL protein samples were developed in 5 mg/mL 2,2'-azino-di(3-ethylbenzthiazoline sulfonate) (ABTS) and 0.01% H2O2. Absorbance changes at 405 nm were used to calculate enzyme activity based on the following equation: (∆ A 405 nm / min Test − ∆ A 405 nm / min Blank ) (V total ) (df ) Units / mL enzyme = (36 . 8 ) (V enzyme ) Vtotal = total volume of assay; df = dilution factor; Venzyme = the volume of enzyme solution; 36.8 = millimolar extinction coefficient of oxidized ABTS at 405 nm. Using fluorescence polarization (FP) assay to determine the binding of TG6-FITC or R9-FITC to lipid vesicles or heparan sulfate. A chloroform solution containing 0.01 mmol total lipids (45% egg yolk phosphatidylcholine, 20% phosphatidylethanolamine, 20% sphingomyelin and 15% cholesterol by molar ratio) was placed in a 10 mL round-bottom flask. Chloroform was removed at room temperature by rotary evaporator to form a uniform lipid film.39 The flask was placed under vacuum for an additional 6 h. The dried lipid film was rehydrated with 1 mL PBS buffer (pH 7.4). The suspension was placed in an ice-water bath and sonicated by using a probe-type ultrasonic homogenizer for 10 min. A typical preparation yields a homogeneous solution containing small unilamellar vesicles (SUVs) with average diameter of ~80 nm and polydispersity index (PDI) less than 0.15. A SpectraMax M5 microplate reader was set to FP mode, with an excitation wavelength of 485 nm and an emission

Page 2 of 8

wavelength of 525 nm. The G-factor was standardized by a fluorescein solution which has a FP of 27 mP. TG6-FITC or R9-FITC with a concentration of 100 nM were incubated with various concentrations of SUVs or heparan sulfate for 1 h in dark and measured for FP, with PBS as blank sample, and SUVs or heparan sulfate solutions as controls. EC50 values were determined with Igor Pro 6 software using Hill equation fitting based on three replicates. Cytotoxicity of TG6 on different human cell lines. The cytotoxicity of TG6 was determined by MTT assay. HeLa, A549 and HepG2 cells were maintained in DMEM supplemented with 10% Fetal Bovine Serum (FBS) and 1% Penicillin/Streptomycin at 37 °C under 5% CO2, and seeded onto a 96-well plate at a density of 10,000 cells per well the day before experiments. TG6 was added to each well at the concentrations as denoted, and incubated with cells for 24 h. Then cells were treated with MTT at 0.5 mg/mL for 4 h, lysed with 150 μL DMSO, and the optical densities were read at 570 nm by an iMark microplate reader. EC50 values were determined with Igor Pro 6 software using sigmoidal curve fitting based on three replicates. In vitro cellular uptake assay. Cellular uptake was evaluated by confocal microscopy and fluorescence activated cell sorting (FACS). For fixed cell confocal microscopy, HeLa cells were incubated with R9-FITC, TG3-FITC, or TG6-FITC at a concentration of 1 μM, BSA-FITC or TG6-BSA-FITC at a concentration of 1.5 μM, and LysoTracker Red DND-99 (Invitrogen) at a concentration of 50 nM, all for the mentioned time periods. These cells were then washed with PBS for three times before fixed in 4% formaldehyde. To visualize early endosomes, the sample was incubated with early endosome marker Rabbit anti-EEA1 antibody (Abcam), and labeled with Alexa Fluor 594 conjugated Goat anti-Rabbit IgG (Abcam). After incubation, all the samples were mounted with Prolong Gold Antifade Reagent with DAPI (Invitrogen). Mounted samples were observed on a LSM 710 confocal microscope. Live-cell confocal microscopy experiments were carried out similar to fixed cell confocal microscopy, except that HeLa cells were observed directly in glass bottom dishes (MatTek) after incubation followed by thorough wash with PBS buffer. For FACS experiments, HeLa cells were seeded onto a 3.5 cm plate at a density of 200,000 cells per plate. Prior to the assay, HeLa cells were washed with PBS buffer and replaced with fresh DMEM, with or without 10% Fetal Bovine Serum (FBS) in the medium. R9-FITC or TG6-FITC were dissolved in PBS and then added to the cell culture medium at a final concentration of 1 μM. BSA-FITC and TG6-BSA-FITC in PBS solution were added to the cell culture medium at a final concentration of 1.5 μM except where otherwise mentioned. Cells were then washed with PBS for three times, and then harvested by trypsin digestion. After centrifuge down, cells were re-suspended in FACS buffer (1% FBS/1 mM EDTA in PBS) and analyzed by a CytoFLEX Flow Cytometer (Beckman Coulter). Inhibition of cellular uptake of TG6-BSA-FITC. FACS was used to analyze the uptake inhibition on living

ACS Paragon Plus Environment

Page 3 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Scheme 1. (A) Chemical structures of TG6, TG6-FITC and TG6-protein conjugates. (B) A facile synthetic route to TGn molecules. HeLa cells. Inhibitors including Amiloride (50 μM), NaN3 (10 mM), Chlorpromazine (CPZ, 20 μg/mL), βcyclodextrin (β-CD, 5 mM) were added into cells culture medium 30 min before adding 1.5 μM TG6-BSA-FITC. After 4 h incubation, cells were washed with PBS for three times, harvested by trypsin digestion, re-suspended in FACS buffer, and analyzed by a CytoFLEX Flow Cytometer. Visualization of TG6-HRP uptake. 3,3',5,5'tetramethyl-benzidine (TMB) is a membrane permeable HRP substrate, which was used to visualize HRP or TG6HRP inside living cells. HeLa cells were seeded onto a 96 well plate at a density of 10,000 cells per well, and incubated with 1 μM TG6-HRP or HRP for 4 h. Cells were then washed with PBS and developed with 300 μL 1-step UltraTMB substrate solution (Pierce) in dark for 0.5 h, and observed under light microscope. Quantification of intracellular TG6-HRP concentration. HeLa cells were seeded onto 3.5 cm plates at a density of 200,000 cells per plate, and treated with TG6HRP or HRP at denoted concentrations for 4 h. Cells were washed with PBS for three times, harvested by trypsin digestion, and lysed in 200 μL 1% Triton X-100 in PBS. TG6-HRP and HRP solutions were also diluted in 1% Triton X-100 in PBS as standards. All samples were then developed in 5 mg/mL ABTS and 0.01% H2O2. Absorbance at 405 nm were monitored by a microplate reader for 10 min. TG6-HRP and HRP concentrations in HeLa cell lysate were calculated based on fitting the slopes of absorbance change of cell lysate samples to TG6-HRP and HRP standards, in triplicates. Intracellular protein total concentrations were determined by using a Micro BCA Protein Assay Kit (Sangon Biotech) according to the manufacture's protocol. The intracellular concentration of TG6HRP or HRP were reported as "ng HRP per mg total proteins in cell".

RESULTS AND DISCUSSION Structure design of the miniature transporter. We chose 1,3,5-triazine as a rigid planar core, and attached four Gdms to the core via flexible alkyl linkers to give the

TGn molecules (Scheme 1A). The space distribution of four Gdms is confined to two cone-like areas around the rigid triazine plane, which would enhance the Gdm-lipid membrane interaction.25 The choice of alkyl chains as linkers would offer structural flexibility and hydrophobicity, which are favorable for lipid membrane binding.40 Comparing with other Gdm-rich synthetic analogs,33-35, 37, 40 TGn has a simpler structure and a lower molecular weight, and is facilely synthesized with a high yield through sequential substitution of the chlorides on cyanuric chloride by secondary amines bearing Gdms (Scheme 1B, see SI for synthesis details and characterization of compounds). Cytosolic delivery of small cargoes by TG6. Nonaarginine (R9) is one of the most potent RPPs,33 which was chosen as the reference for evaluating the transduction activity of TGn on HeLa cells. Both TGn and R9 were conjugated with fluorescein isothiocyanate (FITC) to give TGn-FITC and R9-FITC (Scheme 1, S1-S3). The cellular uptake of TG3-FITC was poor (Figure S1), while the cellular uptake of both R9-FITC and TG6-FITC were high (Figure 1A and S1), even at a low feeding concentration of 1 μM. The different cellular uptake between TG3-FITC and TG6-FITC indicates the effect of alkyl linkers, consistent with the results of Gdm-rich peptoids with different alkyl linkers.40 To rule out the possibility that the observed cellular uptake could be an artifact caused by chemical fixation of cells prior to microcopy imaging,41 confocal experiments and flow cytometry assay (FACS) on living HeLa cells were performed. Inside living HeLa cells (Figure 1A), R9-FITC shows mainly punctate fluorescence, while TG6-FITC shows mostly diffuse fluorescence with some punctate fluorescence. As measured by FACS (Figure 1B), the cellular uptake of R9-FITC is very high at the absence of serum, but is greatly depressed at the presence of 10% serum in the cell culture medium, consistent with literature results.33 In contrast, the cellular uptake of TG6FITC is not affected by serum. At the presence of 10% serum, HeLa cells treated with TG6-FITC show a 6-fold higher cellular uptake than the R9-FITC treated HeLa cells, indicating the advantage of TG6 as a serum-stable

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. (A) Live-cell confocal microscopy images and (B) FACS analysis of HeLa cells after incubation with 1 μM TG6-FITC or R9-FITC for 4 h. There was 10% serum in the cell culture medium for (A). Scale bar is 20 μm. (C,D) Binding of TG6-FITC and R9-FITC to SUVs (C) and heparan sulfate (D), evaluated by FP assay. transporter. Therefore, all the following cell experiments were performed with 10% serum in the cell culture medium. Interaction between TG6 and lipid membrane. To understand the mechanism of cellular uptake of R9 and TG6, we prepared small unilamellar vesicles (SUVs) mimicking the outer membrane of mammalian cells and tested their binding to TG6-FITC and R9-FITC (each at 100 nM) by a fluorescence polarization (FP) assay.42 Based on the binding curves shown in Figure 1C, TG6-FITC bound to the neutral SUVs tightly, with an EC50 value (lipid concentration at which half of TG6-FITC is bound) of 170 μM, while R9-FITC showed a weak binding to the SUVs (EC50 >10 mM), consistent with previous reports.26 The affinity of TG6-FITC to neutral lipid membrane is attributed to its amphiphilic nature.26 We also tested the the binding of TG6-FITC and R9-FITC (each at 100 nM) to heparan sulfate, which is located on cell surface and considered as the primary binding target of RPPs.43 Based on the binding curves shown in Figure 1D, R9-FITC bound to heparan sulfate tightly, with an EC50 value of 250 nM (heparan sulfate concentration at which half of R9-FITC is bound), while TG6-FITC showed no detectable binding to heparan sulfate, which is probably due to the low positive charge (+3) of TG6-FITC, lacking electrostatic interaction with heparan sulfate. The behavior of R9-FITC is consistent with previous reports that R9 binds tightly with cell surface proteoglycans but only weakly with membrane lipids,26, 42 while TG6-FITC behaves more like an amphiphile binding lipid membrane.44 The difference in cellbinding behaviors could lead to different pathways for cellular uptake. TG6 shows minimal cytotoxicity on human cell lines. Previous studies have shown the considerable cytotoxicity of RPPs.31-32 For example, EC50 of oligoarginine peptides is ~ 100 μM for 2 h incubation with C2C12 mouse myoblasts,31 or ~ 10 μM for 7 h incubation with A549 or HeLa cells.32 In contrast, incubation of HeLa cells with

Page 4 of 8

Figure 2. (A,B) MALDI-TOF mass analysis of BSA, BSAFITC and TG6-BSA-FITC (A), and HRP and TG6-HRP (B), showing the average MW of each sample. (C) CD spectra of BSA, BSA-FITC and TG6-BSA-FITC. (D-F) CD spectra of HRP and TG6-HRP in the far-UV region (D), and in the near-UV region (E) and visible region (F). These CD spectra indicate minimal protein structure change after modification. TG6 at 100 μM for 24 h led to less than 10% loss in cell viability (Figure S2). MTT cytotoxicity studies of TG6 with three human cell lines derived from cervix (HeLa), lung (A549), and liver (HepG2) show that the EC50 values of TG6 are in the range of 250-330 μM, indicating the minimal cytotoxicity of TG6 (Table S1), which is much lower than that of oligoarginine peptides. 31-32 Chemical modification of proteins by TG6. To probe the capability of TG6 for protein delivery, we use bovine serum albumin (BSA, a 67-kDa globular protein) and horseradish peroxidase (HRP, a 44-kDa enzyme) as model proteins. BSA was first labeled by FITC to give BSA-FITC for visualization. Both BSA-FITC and HRP were conjugated with TG6 via convenient amine-carboxylate coupling reaction to form TG6-BSA-FITC or TG6-HRP. Based on the ma ss ana ly s is , th e ave ra ge mola r ra tio o f BSA:FITC:TG6 is 1:0.8:1.9 in TG6-BSA-FITC (Figure 2A); the average molar ratio of HRP:TG6 is 1:2.7 in TG6-HRP (Figure 2B). Circular dichroism spectra (CD) showed negligible change of the BSA protein secondary structure after modifications (Figure 2C). Comparison of CD spectra in the 200-250 nm region (for secondary structure) and 250-350 nm region (for tertiary structure) indicates negligible structure change of the HRP protein after TG6 modification (Figure 2D,E).45 The CD spectra in 350-450 nm region is sensitive to the structure change of the Heme pocket, which is the active center of HRP.45 The coincidence of two CD spectra in this region suggests that the structure of the Heme pocket was not affected by TG6 modification (Figure 2F). In fact, the enzyme activity of

ACS Paragon Plus Environment

Page 5 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Table 1. Characterization of modified proteins in PBS buffer. Protein

Hydrodynamic diameter [a] (nm)

Zeta poten[a] tial (mV)

Enzyme activity [a] (U/mg)

BSA

8.1 ± 0.1

-9.8 ± 0.6

-

TG6-BSAFITC

10.1 ± 0.5

-8.6 ± 0.9

-

HRP

6.9 ± 0.1

-7.0 ± 0.8

571 ± 8

TG6-HRP

7.4 ± 0.1

-6.2 ± 0.9

691 ± 23

[a] Values were expressed as mean ± SD of three replicates.

TG6-HRP was measured to be a little higher than the native HRP sample (Table 1). The hydrodynamic diameter and zeta potential of these proteins in PBS buffer (pH = 7.4) were measured by dynamic light scattering (Table 1). The hydrodynamic diameter of modified proteins was slightly bigger than that of native proteins, which indicates that TG6-BSA-FITC and TG6-HRP remain as monomeric proteins in PBS solution.46 The zeta potentials of TG6-protein conjugates remain negative (-8.6 mV for TG6-BSA-FITC, -6.2 mV for TG6-HRP), due to the low modification level (1-3 TG6 per protein). The negative surface potential distinguishes TG6-protein conjugates from most of other transporter-protein conjugates or complexes, which rely on their positive surface potential for cell membrane binding and entry.11, 15, 20 Cellular uptake of TG6-protein conjugates. FACS analysis of living cells provided a quantitative comparison of cellular uptake of TG6-BSA-FITC and BSA-FITC (Figure 3A-C). To remove any protein bound to cell surface, HeLa cells were thoroughly washed by PBS and then treated by trypsin digestion.47 Based on the fluorescence intensity, the cellular uptake of TG6-BSA-FITC was ~50 times higher than that of BSA-FITC (Figure 3A). Similar results were observed for living A549 and HepG2 cells, which showed ~160 (A549) or ~60 (HepG2) times higher uptake of TG6BSA-FITC than that of BSA-FITC (Figure S3). The cellular uptake of TG6-BSA-FITC also increased with the feeding concentration of TG6-BSA-FITC (Figure 3B) and the incubation time (Figure 3C). The feeding concentration of TG6-BSA-FITC (0.15-1.5 μM) is lower than that of proteinRPPs conjugates, which is typically in the range of 5-50 μM,25-27 sometimes as high as 150 μM.29 In addition, the delivery of TG6-BSA-FITC is compatible with serum, while RPPs and protein-RPPs conjugates typically require the absence of serum in the cell culture medium.25-27, 29 The comparison clearly indicates that TG6 is a highly efficient and serum-stable transporter for protein delivery. Uptake mechanism of TG6-protein conjugates. To probe the uptake mechanism of TG6-BSA-FITC, four cell uptake inhibitors were applied (Figure 3D).15 Both chlorpromazine (CPZ) and amiloride inhibit macropinocytosis, but barely restrained the uptake of TG6-BSA-FITC. Similarly, β-cyclodextrin (β-CD) inhibits the caveolaemediated endocytosis, but showed negligibly effect on the uptake of TG6-BSA-FITC. However, ATP depletion by NaN347 reduced the uptake of TG6-BSA-FITC to 65%, in-

Figure 3. (A-C) FACS analysis of live HeLa cells after incubation with 1.5 μM BSA-FITC or TG6-BSA-FITC for 4 h (A), with different concentrations of TG6-BSA-FITC for 4 h (B), and with 1.5 μM TG6-BSA-FITC for different time periods (C). (D) Average fluorescence intensity of live HeLa cells after 4-h incubation with 1.5 μM TG6-BSAFITC at the presence of four cellular uptake inhibitors. Star: p