A fusogenic peptide from a sea urchin fertilization protein promotes intracellular delivery of biomacromolecules by facilitating endosomal escape Keisuke Niikura, Kenichi Horisawa, Nobuhide Doi PII: DOI: Reference:
S0168-3659(15)00627-6 doi: 10.1016/j.jconrel.2015.06.020 COREL 7725
To appear in:
Journal of Controlled Release
Received date: Revised date: Accepted date:
14 November 2014 12 June 2015 15 June 2015
Please cite this article as: Keisuke Niikura, Kenichi Horisawa, Nobuhide Doi, A fusogenic peptide from a sea urchin fertilization protein promotes intracellular delivery of biomacromolecules by facilitating endosomal escape, Journal of Controlled Release (2015), doi: 10.1016/j.jconrel.2015.06.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
SC RI
Keisuke Niikura, Kenichi Horisawa1, Nobuhide Doi*
PT
A fusogenic peptide from a sea urchin fertilization protein promotes intracellular delivery of biomacromolecules by facilitating endosomal escape
NU
Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Yokohama 223-8522, Japan *Corresponding Author. Tel.: +81 45 566 1772; fax: +81 45 566 1440. E-mail address: [email protected] 1 Present address: Division of Organogenesis and Regeneration, Medical Institute of Bioregulation, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan
MA
Abstract
The low efficiency of endosomal escape has been considered a bottleneck for the cytosolic delivery of biomacromolecules such as proteins and DNA. Although
ED
fusogenic peptides (FPs) such as HA2 have been employed to improve the intracellular delivery of biomacromolecules, the FPs studied thus far are not adequately efficient in enabling endosomal escape; therefore, novel FPs with higher activity are required. In
PT
this context, we focused on FPs derived from a sea urchin fertilization protein, bindin, which is involved in gamete recognition (B18, residues 103-120 and B55, residues
CE
83-137 of mature bindin). We show that enhanced green fluorescent protein (EGFP)-fused B55 peptide binds to plasma membranes more strongly than EGFP-B18
AC
and promotes the intracellular delivery of dextrans, which were co-administered using the trans method in a pH-dependent manner without affecting cell viability and proliferation, whereas conventional EGFP-HA2 did not affect dextran internalization. Furthermore, EGFP-B55 promoted the intracellular delivery of biomacromolecules such as antibodies, ribonuclease and plasmidic DNA using the trans method. Because the promotion of intracellular delivery by EGFP-B55 was suppressed by endocytosis inhibitors, EGFP-B55 is considered to have facilitated the endosomal escape of co-administered cargos. These results suggested that an FP that promotes the intracellular delivery of a variety of biomacromolecules with no detectable cytotoxicity should be useful for the cytosolic delivery of membrane-impermeable molecules for biomedical and biotechnological applications. Keywords: bindin; endocytosis; IgG; ribonuclease; gene delivery; quantitative imaging 1
ACCEPTED MANUSCRIPT
1. Introduction
PT
Although biomacromolecules, such as peptides, enzymes, antibodies and oligonucleotides, have advantages with respect to their unique functions and target
SC RI
specificity compared with small-molecule compounds, biomacromolecules need to be delivered into the cytosol or nucleus for their therapeutic application to directly target intracellular molecules [1-3]. Cell-penetrating peptides (CPPs) have so far been employed for the intracellular delivery of various biomacromolecules in the fields of
NU
molecular imaging, nucleic acid delivery, peptide/protein delivery and therapeutics [4,5]. One of the most studied CPPs is an HIV-derived TAT peptide, which is arginine-rich
MA
and interacts with the cell surface, resulting in penetration into cells through various pathways [6-9]. However, several drawbacks were noted for CPP-dependent delivery. First, non-specific penetrations of CPPs may cause instability in vivo until they reach
ED
their targets or cause undesirable side effects because the penetration of cationic CPPs into cells primarily depends on the electrostatic interaction between the cationic charge
PT
of the peptide and the anionic charge of the cell surface [9-11]. To solve this issue, Fei et al. designed a pH-sensitive peptide for targeting the acidic tumor microenvironment to reduce non-specific uptake by masking the cationic charge under the neutral pH of
CE
normal cells [12].
Additionally, the low-efficiency of endosomal escape is a bottleneck for
AC
intracellular delivery by CPPs [13,14]. Although improving the efficiency of endosomal escape by CPPs alone is difficult, the ability to facilitate endosomal escape has been discovered in fusogenic peptides (FPs), and some studies have utilized FPs for intracellular delivery [3,15-17]. For example, Wadia et al. combined the TAT peptide and fusogenic HA2 peptide derived from hemagglutinin from influenza virus and demonstrated that dTAT-HA2 efficiently promoted the escape from macropinosomes [16]. FPs have membrane-disrupting activities under acidic conditions, through pH-dependent conformational changes [18], and consequently, they are expected to facilitate endosomal escape at acidic pH. The pH-dependence of FPs is an important characteristic for their application in drug delivery systems (DDS) with cell targeting because it reduces non-specific effects [15]. However, the FPs studied so far are not adequately efficient in facilitating endosomal escape; thus, novel FPs with higher activity are required. 2
ACCEPTED MANUSCRIPT
Sea urchin Strongylocentrotus purpuratus ‘bindin’ is involved in egg-sperm recognition and fusion in fertilization [19-21]. The 18 amino acid residues of the
PT
minimal membrane-binding region of bindin, B18, identified more than two decades ago, were known to possess fusogenic activity via a Zn2+-dependent conformational
SC RI
change [22-24] and membrane-disrupting properties that induce leakage of the liposomal contents [23]. However, the effects of the B18 peptide on mammalian cells and its application in intracellular delivery have not been studied. Therefore, we investigated the effects of the B18 peptide on the plasma membrane of HeLa cells. In
NU
addition, we compared the functions of B18 with those of the B55 peptide, defined as the ‘bindin core region’ including the B18 peptide [21]. We found that an enhanced
MA
green fluorescent protein (EGFP)-fused B55 peptide bound to the plasma membrane and promoted the intracellular delivery of other co-administered biomacromolecules such as dextrans, antibodies, ribonuclease (RNase) and plasmidic DNA (pDNA). Its
ED
function was suppressed by an endocytosis inhibitor, indicating that the B55 peptide facilitates the endosomal escape of these biomacromolecules. Thus, the B55 peptide is
PT
anticipated to be a valuable tool for efficient intracellular delivery.
CE
2. Materials and methods
AC
2.1. Construction of plasmids All oligonucleotides were purchased from Eurofins Genomics (Tokyo, Japan). The primer sequences are indicated in Supplementary Table S1. EGFP and EGFP-B18 genes were amplified from pEGFP-RHA [25] using the primers His-FLAG-EGFP-F and either EGFP-R or EGFP-B18-R, respectively. These PCR products were cloned into a pET20b(+) vector (Merck KGaA, Darmstadt, Germany) at the NdeI and XhoI sites, resulting in pEGFP and pEGFP-B18. B55 (from S. purpuratus bindin gene sequence NM_214518.1), modified HA2 (INF7) [26] and TAT genes were produced by repeated cycles of hybridization of two partially complementary DNAs, B55-F and B55-R, HA2-F and HA2-R, and TAT-F and TAT-R, respectively, at 55°C for 30 sec, followed by extension with Phusion DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA) at 72°C for 1 min, after denaturation at 98°C for 30 sec. The B18-TAT gene was amplified from pEGFP-B18 3
ACCEPTED MANUSCRIPT
using primers B18-F and B18-TAT-R. These genes were cloned into pEGFP-B18 at the BamHI and XhoI sites, resulting in pEGFP-B55, pEGFP-HA2, pEGFP-TAT and
PT
pEGFP-B18-TAT. The DNA sequences of all plasmids were confirmed by Applied
SC RI
Biosystems 3130xl Genetic Analyzer (Life Technologies, Carlsbad, CA, USA). 2.2. Protein expression and purification
All proteins were expressed in Escherichia coli BL21(DE3)-CodonPlus-RIL (Agilent Technologies, Santa Clara, CA, USA). Each expression vector was
NU
transformed into bacterial cells, and one clone of each sample was cultivated in 200 ml of Luria-Bertani (LB) medium containing 100 mg/ml ampicillin at 20ºC for 72 hr. After
MA
cultivation, additional overnight incubation was performed at 4ºC to facilitate EGFP folding. Next, all bacterial cells were collected in two 50-ml centrifuge tubes for each sample and stored at -80ºC until use. TBS (50 mM Tris-HCl, pH 7.6, and 200 mM
ED
NaCl) supplemented with a protease inhibitor cocktail for histidine-tag (Sigma-Aldrich, St. Louis, MO, USA) was added to each tube, followed by sonication. Each sample was
PT
centrifuged (8500 x g, 10 min), and the supernatant was harvested, except for EGFP-B55. For EGFP-B55, TBS with 6 M urea and the protease inhibitor cocktail were added to the resulting pellet from the prior step, and the sample was vortexed at 4ºC for was harvested.
CE
1 hr. The solubilized sample was centrifuged (8500 x g, 10 min), and the supernatant
AC
Each protein was purified using Cosmogel His-Accept (Nacalai Tesque, Kyoto, Japan). Phosphate buffer (PB; 90 mM NaH2PO4, 90 mM Na2HPO4, and 500 mM NaCl) with 0.1% Tween 20 was used for washing the Ni2+ beads, and PB with 300 mM imidazole was used for elution from the Ni2+ beads. EGFP-B55 was purified under denaturing conditions using 6 M urea. After purification, each buffer containing the purified samples was exchanged with PB using Amicon Ultra-4 (30k; Merck Millipore, Billerica, MA, USA). To confirm the pH dependency as described in section 2.4, PB with varying pH values was used. Protein concentrations were measured using a BCA protein assay kit (Thermo Fisher Scientific). The purification of all proteins was confirmed using 12.5% SDS-PAGE and Coomassie brilliant blue (CBB) staining. 2.3. Cell line and cell culture The human cervical cancer cell line HeLa (RIKEN Cell Bank, Ibaraki, Japan) was 4
ACCEPTED MANUSCRIPT
maintained in Dulbecco's modified Eagle's medium (DMEM) (Nacalai Tesque) with 10% (v/v) fetal bovine serum (Nichirei Biosciences Inc., Tokyo, Japan) and 1% (v/v)
PT
penicillin-streptomycin (Life Technologies) and incubated at 37ºC and 5% CO2 in static culture.
SC RI
For the confirmation of pH dependency, three types of DMEM, with pH 6.8, 7.1 and 7.7, were prepared by adding 10% PB with pH values 6.1, 6.7 and 7.1, respectively. The final pH of each medium was measured using an F-52 pH meter (Horiba, Kyoto,
NU
Japan). 2.4. Fluorescent imaging
MA
The HeLa cells were seeded in a glass base dish (AGC Techno Glass Co., Ltd., Shizuoka, Japan) prior to the experiments for 24 hr. The medium of each sample was replaced by DMEM containing 10% PB, 10% (v/v) fetal bovine serum, 1% (v/v)
ED
penicillin-streptomycin, 10 µM of the respective EGFP fusion proteins and 1 µg/ml Hoechst 33342. To study the intracellular delivery, 20 µM dextrans (3, 10 and 40 kDa)
PT
labeled with Texas Red (Thermo Fisher Scientific) and 50 µg/ml Alexa Fluor 568 goat anti-rabbit IgG (H+L) antibody (Life Technologies) or 25 µg/ml anti-γ-tubulin-Cy3 antibody (Sigma-Aldrich) were added to the medium described above. For the
CE
confirmation of endocytosis, 20 µM dynasore (Sigma-Aldrich) or 10 µM EIPA [5-(N-ethyl-N-isopropyl)-amiloride; Sigma-Aldrich], and/or 0.1% DMSO was added.
AC
For the confirmation of the influence of Zn2+, various concentrations of ZnCl2 (Nacalai Tesque) were added to the above medium without PB. After incubation, the cells were washed three times by D-PBS (-) (Nacalai Tesque). Next, the cells were fixed using 4% paraformaldehyde phosphate buffer solution (Nacalai Tesque) for 30 min. After fixation, the cells were washed twice with D-PBS (-) and observed by confocal microscopy (FV1000, Olympus, Tokyo, Japan) as soon as possible. On the other hand, for live-cell imaging, medium of each dish was replaced to DMEM without phenol red (Nacalai Tesque) containing 10% (v/v) fetal bovine serum and 1% (v/v) penicillin-streptomycin and observed by confocal microscopy as soon as possible. Quantitative analysis was performed using FV10-ASW (Olympus), and the relative fluorescent intensities were calculated using the ‘average’ parameters of the ROI defined in each cell. For immunocytochemistry, the HeLa cells were seeded in a glass base dish prior to the experiments for 24 hr and then incubated with 10 µM EGFP or EGFP-B55 for 24 5
ACCEPTED MANUSCRIPT
hr. After incubation, the cells were washed three times using D-PBS (-). Next, the cells were fixed using 4% paraformaldehyde phosphate buffer solution for 30 min and treated
PT
with 0.2% Triton X-100/PBS for 5 min. The cells were washed twice with D-PBS (-) and blocked using 3% BSA/PBS for 30 min. Next, they were treated with
SC RI
anti-γ-tubulin-Cy3 antibody and SlowFade Gold Antifade Mountant with DAPI (Life Technologies) in 3% BSA/PBS for 3 hr at 4ºC. The cells were washed with D-PBS (-) twice and observed using confocal microscopy.
NU
2.5. RNase assays
The HeLa cells (1.0 x 105) were seeded in 12-well dishes (Corning Inc., Corning,
MA
NY, USA) prior to the experiments for 24 hr. The medium of each sample was replaced with DMEM containing 10% PB, 10% (v/v) fetal bovine serum, 1% (v/v) penicillin-streptomycin, 10 µM EGFP or EGFP-B55 and 500 µg/ml RNase A from
ED
bovine pancreas (Nacalai Tesque), and the samples were incubated for 24 hr. After incubation, the RNA from each sample was isolated using TRIzol reagent (Life
PT
Technologies) according to the manufacturer’s protocol. The purified RNA was re-dissolved in Ultra Pure (Life Technologies), and the RNA concentration of each
CE
sample was measured using Nano Drop 1000 (Thermo Fisher Scientific). 2.6. Luciferase gene delivery
AC
The HeLa cells (5.0 x 104) were seeded in 24-well dishes (AGC Techno Glass Co., Ltd.) prior to the experiments for 24 hr. The medium of each sample was replaced with Opti-MEM (Life Technologies) containing 10% PB and 10 µM EGFP or EGFP-B55. The cells were transfected with 0.27 µg of pGL4.13 luc2/SV40 (Promega, Fitchburg, WI, USA) and 0.67 µl of Lipofectamine 2000 (Life Technologies) per well. After a 5-hr incubation, the medium of each sample was replaced with DMEM containing 10% (v/v) fetal bovine serum and 1% (v/v) penicillin-streptomycin and incubated again for 19 hr. The HeLa cells were washed with D-PBS (-), and the luciferase assay was performed according to the manufacturer’s protocol using the Luciferase Assay System (Promega) with a Gene Light luminometer (Microtec Co., Ltd, Chiba, Japan). Relative light units (RLU) for each sample were normalized with respect to their protein concentrations as measured using the BCA protein assay kit.
6
ACCEPTED MANUSCRIPT
2.7. Cell viability and cytotoxicity assays In the WST-1 assays, the HeLa cells (1.0 x 103) were seeded in a 96-well plate
PT
(Thermo Fisher Scientific) prior to the experiments for 24 hr. The medium of each well was replaced with DMEM containing 10% PB, 10% (v/v) fetal bovine serum, 1% (v/v)
SC RI
penicillin-streptomycin and 10 µM of each protein. After incubation for 4, 8, 24, or 48 hr, the medium of each well was replaced with the mixture of 100 µl of DMEM and 10 µl of the cell proliferation reagent WST-1 (Hoffmann-La Roche, Basel, Switzerland), and the cells were incubated for 2 hr. After incubation, the absorbance of each sample
NU
was measured using a Safire microplate reader (Tecan Group Ltd., Männedorf, Switzerland).
MA
In the LDH assays, the cells were incubated as well as the WST-1 assays for 24 hr and the detection of LDH activity was performed according to the manufacturer’s protocol using the CytoTox-ONE homogeneous membrane integrity assay (Promega).
ED
The cytotoxicity of each sample was calculated from the fluorescent intensities of the difference between the samples treated with reagents and the samples without reagents
PT
per the samples from lysed cells with reagents. 2.8. Statistical analysis
AC
CE
All statistical analysis was performed according to Welch’s t-test method.
3. Results
3.1. Construction and expression of EGFP-fusogenic peptide fusion proteins First, we designed and constructed EGFP fusion proteins with FPs derived from the sea urchin S. purpuratus mature bindin protein (B18, 103-120 aa and B55, 83-137 aa) (Fig. 1A). All proteins were expressed in E. coli and purified using Ni2+ affinity chromatography with >90% homogeneity as confirmed by the CBB staining of the SDS-PAGE gel (Fig. 1B). Next, we investigated the localization of purified EGFP-B18 and EGFP-B55 added to the HeLa cells. After a 4-hr incubation of each purified proteins with the HeLa cells, EGFP-B18 and EGFP-B55 were localized to the plasma membranes of the HeLa cells (Fig. 1C). Notably, EGFP-B55 was more strongly bound to the plasma membranes 7
ACCEPTED MANUSCRIPT
than EGFP-B18. This result is consistent with a previous report that the regions 69-91 aa and 119-130 aa of mature bindin were associated with the binding affinity of the
PT
peptide to the lipid bilayer [22]: the B18 peptide does not contain these regions, whereas the B55 peptide includes more than half of these regions, suggesting that the extended
SC RI
peptides of both sides of the B18 peptide contribute to its relatively higher affinity to lipid bilayers.
The low efficiency of the binding of the B18 peptide to the plasma membranes is common to other FPs, and FP-mediated endosomal escape has been confirmed by the
NU
additional fusion of CPPs to increase binding to plasma membranes [16,17]. Thus, we confirmed the function of EGFP-B18 using an additional fusion of a TAT peptide
MA
(Supplementary Fig. S1A). Although EGFP-TAT was primarily localized to endosome-like vesicles, most of the EGFP-B18-TAT was localized to the cytosol (Supplementary Fig. S1B). Quantitative image analysis revealed that the fluorescent
ED
intensity of intracellular EGFP-B18-TAT was approximately ten times higher than that of EGFP-TAT (Supplementary Fig. S1C). These results suggest that the B18 peptide
PT
facilitates the endosomal escape of EGFP-B18-TAT itself. Although the conjugation of a CPP-FP fusion peptide with a large protein, such as mCherry [27] or an antibody [28], sometimes reduces the endosomolytic activity of the CPP-FP peptide, the EGFP-fused
CE
B18-TAT peptide retained sufficient activity in facilitating endosomal escape. Therefore, the B18-TAT peptide is possibly suitable for the intracellular delivery of proteins
AC
directly conjugated to it.
8
AC
CE
PT
ED
MA
NU
SC RI
PT
ACCEPTED MANUSCRIPT
Fig. 1. Construction and expression of EGFP-FP fusion proteins. (A) Amino acid sequences of EGFP-fused fusogenic peptides (FPs). TCS; thrombin cleavage site. ‘GS’ dipeptide between EGFP and each additional peptide corresponds to a BamHI site. In EGFP-B18, there is an additional linker ‘PYD’ between EGFP and the B18 peptide (red), which is derived from parental bindin. (B) SDS-PAGE analysis of EGFP fusion proteins. EGFP fusion proteins were expressed in E. coli BL21(DE3) and purified using the Ni2+ beads column. The bands were detected using CBB staining. The molecular weights of EGFP, EGFP-B18, EGFP-B55 and EGFP-HA2 are 29.5, 32.1, 35.7 and 32.2 kDa, respectively. (C) Confocal microscopic images of the HeLa cells co-treated with 10 µM of the respective EGFP fusion protein (green) and 1 µg/ml Hoechst 33342 (blue) for 4 hr.
9
ACCEPTED MANUSCRIPT
3.2. EGFP-B55 promoted the intracellular delivery of dextrans Recently, Salomone et al. reported that a hybrid peptide of two antimicrobial
PT
peptides and a CPP from the HIV-1 Tat protein, CM18-TAT, had membrane-disruptive properties and enhanced the intracellular delivery of dextrans of various sizes by
SC RI
promoting endosomal escape [29]. Therefore, we investigated whether EGFP-B18-TAT and EGFP-B55 could enhance the plasma membrane permeability of dextrans. When EGFP-B55 and Texas Red-labeled dextrans were co-incubated with the HeLa cells, the dextrans were localized not only to endosome-like vesicles but also to both the
NU
cytoplasm and nucleus (Fig. 2A). The fluorescent intensity of the dextrans was significantly higher after co-incubation with EGFP-B55 than with EGFP (Fig. 2B).
MA
Histograms of each sample showed monodispersed peaks, indicating that the intracellular delivery of dextrans was promoted by EGFP-B55 in all incubated cells (Fig. 2C). This localization of EGFP-B55 and dextrans in fixed cells was similarly observed
ED
in live cells (Supplementary Fig. S2). In addition, the time-course experiments of the localization of EGFP-B55 and dextrans revealed that the amount of binding of
PT
EGFP-B55 to the cell membranes was gradually increased from 4 hr through 24 hr, while the intracellular delivery of dextrans in the presence of EGFP-B55 was rapidly increased between 8 and 24 hr (Supplementary Fig. S3). These results suggested that
CE
EGFP-B55 significantly promoted the intracellular delivery and endosomal escape of dextrans but that EGFP-HA2, a conventional FP derived from influenza virus
AC
hemagglutinin, did not affect dextran internalization. Because the activity of EGFP-B18-TAT was markedly less than that of EGFP-B55 (Supplementary Fig. S1D), the following analyses were performed on EGFP-B55.
10
AC
CE
PT
ED
MA
NU
SC RI
PT
ACCEPTED MANUSCRIPT
Fig. 2. Intracellular delivery of dextran 3000 in the presence of EGFP-FP fusion proteins. (A) Confocal microscopic images of the HeLa cells co-treated with 10 µM of the respective EGFP fusion protein (green), 60 µg/ml Texas Red-conjugated dextran 3000 (red) and 1 µg/ml Hoechst 33342 (blue) for 24 hr. Every medium had a pH of 7.1. The right panels correspond to the inset in left panels. (B) Relative fluorescent intensities of Texas Red per cell in the presence of each EGFP fusion protein. The intensities were normalized with respect to the average of the samples treated with EGFP. N = 59-75. Mean ± SD. * p
S0168-3659(15)00627-6 doi: 10.1016/j.jconrel.2015.06.020 COREL 7725
To appear in:
Journal of Controlled Release
Received date: Revised date: Accepted date:
14 November 2014 12 June 2015 15 June 2015
Please cite this article as: Keisuke Niikura, Kenichi Horisawa, Nobuhide Doi, A fusogenic peptide from a sea urchin fertilization protein promotes intracellular delivery of biomacromolecules by facilitating endosomal escape, Journal of Controlled Release (2015), doi: 10.1016/j.jconrel.2015.06.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
SC RI
Keisuke Niikura, Kenichi Horisawa1, Nobuhide Doi*
PT
A fusogenic peptide from a sea urchin fertilization protein promotes intracellular delivery of biomacromolecules by facilitating endosomal escape
NU
Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Yokohama 223-8522, Japan *Corresponding Author. Tel.: +81 45 566 1772; fax: +81 45 566 1440. E-mail address: [email protected] 1 Present address: Division of Organogenesis and Regeneration, Medical Institute of Bioregulation, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan
MA
Abstract
The low efficiency of endosomal escape has been considered a bottleneck for the cytosolic delivery of biomacromolecules such as proteins and DNA. Although
ED
fusogenic peptides (FPs) such as HA2 have been employed to improve the intracellular delivery of biomacromolecules, the FPs studied thus far are not adequately efficient in enabling endosomal escape; therefore, novel FPs with higher activity are required. In
PT
this context, we focused on FPs derived from a sea urchin fertilization protein, bindin, which is involved in gamete recognition (B18, residues 103-120 and B55, residues
CE
83-137 of mature bindin). We show that enhanced green fluorescent protein (EGFP)-fused B55 peptide binds to plasma membranes more strongly than EGFP-B18
AC
and promotes the intracellular delivery of dextrans, which were co-administered using the trans method in a pH-dependent manner without affecting cell viability and proliferation, whereas conventional EGFP-HA2 did not affect dextran internalization. Furthermore, EGFP-B55 promoted the intracellular delivery of biomacromolecules such as antibodies, ribonuclease and plasmidic DNA using the trans method. Because the promotion of intracellular delivery by EGFP-B55 was suppressed by endocytosis inhibitors, EGFP-B55 is considered to have facilitated the endosomal escape of co-administered cargos. These results suggested that an FP that promotes the intracellular delivery of a variety of biomacromolecules with no detectable cytotoxicity should be useful for the cytosolic delivery of membrane-impermeable molecules for biomedical and biotechnological applications. Keywords: bindin; endocytosis; IgG; ribonuclease; gene delivery; quantitative imaging 1
ACCEPTED MANUSCRIPT
1. Introduction
PT
Although biomacromolecules, such as peptides, enzymes, antibodies and oligonucleotides, have advantages with respect to their unique functions and target
SC RI
specificity compared with small-molecule compounds, biomacromolecules need to be delivered into the cytosol or nucleus for their therapeutic application to directly target intracellular molecules [1-3]. Cell-penetrating peptides (CPPs) have so far been employed for the intracellular delivery of various biomacromolecules in the fields of
NU
molecular imaging, nucleic acid delivery, peptide/protein delivery and therapeutics [4,5]. One of the most studied CPPs is an HIV-derived TAT peptide, which is arginine-rich
MA
and interacts with the cell surface, resulting in penetration into cells through various pathways [6-9]. However, several drawbacks were noted for CPP-dependent delivery. First, non-specific penetrations of CPPs may cause instability in vivo until they reach
ED
their targets or cause undesirable side effects because the penetration of cationic CPPs into cells primarily depends on the electrostatic interaction between the cationic charge
PT
of the peptide and the anionic charge of the cell surface [9-11]. To solve this issue, Fei et al. designed a pH-sensitive peptide for targeting the acidic tumor microenvironment to reduce non-specific uptake by masking the cationic charge under the neutral pH of
CE
normal cells [12].
Additionally, the low-efficiency of endosomal escape is a bottleneck for
AC
intracellular delivery by CPPs [13,14]. Although improving the efficiency of endosomal escape by CPPs alone is difficult, the ability to facilitate endosomal escape has been discovered in fusogenic peptides (FPs), and some studies have utilized FPs for intracellular delivery [3,15-17]. For example, Wadia et al. combined the TAT peptide and fusogenic HA2 peptide derived from hemagglutinin from influenza virus and demonstrated that dTAT-HA2 efficiently promoted the escape from macropinosomes [16]. FPs have membrane-disrupting activities under acidic conditions, through pH-dependent conformational changes [18], and consequently, they are expected to facilitate endosomal escape at acidic pH. The pH-dependence of FPs is an important characteristic for their application in drug delivery systems (DDS) with cell targeting because it reduces non-specific effects [15]. However, the FPs studied so far are not adequately efficient in facilitating endosomal escape; thus, novel FPs with higher activity are required. 2
ACCEPTED MANUSCRIPT
Sea urchin Strongylocentrotus purpuratus ‘bindin’ is involved in egg-sperm recognition and fusion in fertilization [19-21]. The 18 amino acid residues of the
PT
minimal membrane-binding region of bindin, B18, identified more than two decades ago, were known to possess fusogenic activity via a Zn2+-dependent conformational
SC RI
change [22-24] and membrane-disrupting properties that induce leakage of the liposomal contents [23]. However, the effects of the B18 peptide on mammalian cells and its application in intracellular delivery have not been studied. Therefore, we investigated the effects of the B18 peptide on the plasma membrane of HeLa cells. In
NU
addition, we compared the functions of B18 with those of the B55 peptide, defined as the ‘bindin core region’ including the B18 peptide [21]. We found that an enhanced
MA
green fluorescent protein (EGFP)-fused B55 peptide bound to the plasma membrane and promoted the intracellular delivery of other co-administered biomacromolecules such as dextrans, antibodies, ribonuclease (RNase) and plasmidic DNA (pDNA). Its
ED
function was suppressed by an endocytosis inhibitor, indicating that the B55 peptide facilitates the endosomal escape of these biomacromolecules. Thus, the B55 peptide is
PT
anticipated to be a valuable tool for efficient intracellular delivery.
CE
2. Materials and methods
AC
2.1. Construction of plasmids All oligonucleotides were purchased from Eurofins Genomics (Tokyo, Japan). The primer sequences are indicated in Supplementary Table S1. EGFP and EGFP-B18 genes were amplified from pEGFP-RHA [25] using the primers His-FLAG-EGFP-F and either EGFP-R or EGFP-B18-R, respectively. These PCR products were cloned into a pET20b(+) vector (Merck KGaA, Darmstadt, Germany) at the NdeI and XhoI sites, resulting in pEGFP and pEGFP-B18. B55 (from S. purpuratus bindin gene sequence NM_214518.1), modified HA2 (INF7) [26] and TAT genes were produced by repeated cycles of hybridization of two partially complementary DNAs, B55-F and B55-R, HA2-F and HA2-R, and TAT-F and TAT-R, respectively, at 55°C for 30 sec, followed by extension with Phusion DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA) at 72°C for 1 min, after denaturation at 98°C for 30 sec. The B18-TAT gene was amplified from pEGFP-B18 3
ACCEPTED MANUSCRIPT
using primers B18-F and B18-TAT-R. These genes were cloned into pEGFP-B18 at the BamHI and XhoI sites, resulting in pEGFP-B55, pEGFP-HA2, pEGFP-TAT and
PT
pEGFP-B18-TAT. The DNA sequences of all plasmids were confirmed by Applied
SC RI
Biosystems 3130xl Genetic Analyzer (Life Technologies, Carlsbad, CA, USA). 2.2. Protein expression and purification
All proteins were expressed in Escherichia coli BL21(DE3)-CodonPlus-RIL (Agilent Technologies, Santa Clara, CA, USA). Each expression vector was
NU
transformed into bacterial cells, and one clone of each sample was cultivated in 200 ml of Luria-Bertani (LB) medium containing 100 mg/ml ampicillin at 20ºC for 72 hr. After
MA
cultivation, additional overnight incubation was performed at 4ºC to facilitate EGFP folding. Next, all bacterial cells were collected in two 50-ml centrifuge tubes for each sample and stored at -80ºC until use. TBS (50 mM Tris-HCl, pH 7.6, and 200 mM
ED
NaCl) supplemented with a protease inhibitor cocktail for histidine-tag (Sigma-Aldrich, St. Louis, MO, USA) was added to each tube, followed by sonication. Each sample was
PT
centrifuged (8500 x g, 10 min), and the supernatant was harvested, except for EGFP-B55. For EGFP-B55, TBS with 6 M urea and the protease inhibitor cocktail were added to the resulting pellet from the prior step, and the sample was vortexed at 4ºC for was harvested.
CE
1 hr. The solubilized sample was centrifuged (8500 x g, 10 min), and the supernatant
AC
Each protein was purified using Cosmogel His-Accept (Nacalai Tesque, Kyoto, Japan). Phosphate buffer (PB; 90 mM NaH2PO4, 90 mM Na2HPO4, and 500 mM NaCl) with 0.1% Tween 20 was used for washing the Ni2+ beads, and PB with 300 mM imidazole was used for elution from the Ni2+ beads. EGFP-B55 was purified under denaturing conditions using 6 M urea. After purification, each buffer containing the purified samples was exchanged with PB using Amicon Ultra-4 (30k; Merck Millipore, Billerica, MA, USA). To confirm the pH dependency as described in section 2.4, PB with varying pH values was used. Protein concentrations were measured using a BCA protein assay kit (Thermo Fisher Scientific). The purification of all proteins was confirmed using 12.5% SDS-PAGE and Coomassie brilliant blue (CBB) staining. 2.3. Cell line and cell culture The human cervical cancer cell line HeLa (RIKEN Cell Bank, Ibaraki, Japan) was 4
ACCEPTED MANUSCRIPT
maintained in Dulbecco's modified Eagle's medium (DMEM) (Nacalai Tesque) with 10% (v/v) fetal bovine serum (Nichirei Biosciences Inc., Tokyo, Japan) and 1% (v/v)
PT
penicillin-streptomycin (Life Technologies) and incubated at 37ºC and 5% CO2 in static culture.
SC RI
For the confirmation of pH dependency, three types of DMEM, with pH 6.8, 7.1 and 7.7, were prepared by adding 10% PB with pH values 6.1, 6.7 and 7.1, respectively. The final pH of each medium was measured using an F-52 pH meter (Horiba, Kyoto,
NU
Japan). 2.4. Fluorescent imaging
MA
The HeLa cells were seeded in a glass base dish (AGC Techno Glass Co., Ltd., Shizuoka, Japan) prior to the experiments for 24 hr. The medium of each sample was replaced by DMEM containing 10% PB, 10% (v/v) fetal bovine serum, 1% (v/v)
ED
penicillin-streptomycin, 10 µM of the respective EGFP fusion proteins and 1 µg/ml Hoechst 33342. To study the intracellular delivery, 20 µM dextrans (3, 10 and 40 kDa)
PT
labeled with Texas Red (Thermo Fisher Scientific) and 50 µg/ml Alexa Fluor 568 goat anti-rabbit IgG (H+L) antibody (Life Technologies) or 25 µg/ml anti-γ-tubulin-Cy3 antibody (Sigma-Aldrich) were added to the medium described above. For the
CE
confirmation of endocytosis, 20 µM dynasore (Sigma-Aldrich) or 10 µM EIPA [5-(N-ethyl-N-isopropyl)-amiloride; Sigma-Aldrich], and/or 0.1% DMSO was added.
AC
For the confirmation of the influence of Zn2+, various concentrations of ZnCl2 (Nacalai Tesque) were added to the above medium without PB. After incubation, the cells were washed three times by D-PBS (-) (Nacalai Tesque). Next, the cells were fixed using 4% paraformaldehyde phosphate buffer solution (Nacalai Tesque) for 30 min. After fixation, the cells were washed twice with D-PBS (-) and observed by confocal microscopy (FV1000, Olympus, Tokyo, Japan) as soon as possible. On the other hand, for live-cell imaging, medium of each dish was replaced to DMEM without phenol red (Nacalai Tesque) containing 10% (v/v) fetal bovine serum and 1% (v/v) penicillin-streptomycin and observed by confocal microscopy as soon as possible. Quantitative analysis was performed using FV10-ASW (Olympus), and the relative fluorescent intensities were calculated using the ‘average’ parameters of the ROI defined in each cell. For immunocytochemistry, the HeLa cells were seeded in a glass base dish prior to the experiments for 24 hr and then incubated with 10 µM EGFP or EGFP-B55 for 24 5
ACCEPTED MANUSCRIPT
hr. After incubation, the cells were washed three times using D-PBS (-). Next, the cells were fixed using 4% paraformaldehyde phosphate buffer solution for 30 min and treated
PT
with 0.2% Triton X-100/PBS for 5 min. The cells were washed twice with D-PBS (-) and blocked using 3% BSA/PBS for 30 min. Next, they were treated with
SC RI
anti-γ-tubulin-Cy3 antibody and SlowFade Gold Antifade Mountant with DAPI (Life Technologies) in 3% BSA/PBS for 3 hr at 4ºC. The cells were washed with D-PBS (-) twice and observed using confocal microscopy.
NU
2.5. RNase assays
The HeLa cells (1.0 x 105) were seeded in 12-well dishes (Corning Inc., Corning,
MA
NY, USA) prior to the experiments for 24 hr. The medium of each sample was replaced with DMEM containing 10% PB, 10% (v/v) fetal bovine serum, 1% (v/v) penicillin-streptomycin, 10 µM EGFP or EGFP-B55 and 500 µg/ml RNase A from
ED
bovine pancreas (Nacalai Tesque), and the samples were incubated for 24 hr. After incubation, the RNA from each sample was isolated using TRIzol reagent (Life
PT
Technologies) according to the manufacturer’s protocol. The purified RNA was re-dissolved in Ultra Pure (Life Technologies), and the RNA concentration of each
CE
sample was measured using Nano Drop 1000 (Thermo Fisher Scientific). 2.6. Luciferase gene delivery
AC
The HeLa cells (5.0 x 104) were seeded in 24-well dishes (AGC Techno Glass Co., Ltd.) prior to the experiments for 24 hr. The medium of each sample was replaced with Opti-MEM (Life Technologies) containing 10% PB and 10 µM EGFP or EGFP-B55. The cells were transfected with 0.27 µg of pGL4.13 luc2/SV40 (Promega, Fitchburg, WI, USA) and 0.67 µl of Lipofectamine 2000 (Life Technologies) per well. After a 5-hr incubation, the medium of each sample was replaced with DMEM containing 10% (v/v) fetal bovine serum and 1% (v/v) penicillin-streptomycin and incubated again for 19 hr. The HeLa cells were washed with D-PBS (-), and the luciferase assay was performed according to the manufacturer’s protocol using the Luciferase Assay System (Promega) with a Gene Light luminometer (Microtec Co., Ltd, Chiba, Japan). Relative light units (RLU) for each sample were normalized with respect to their protein concentrations as measured using the BCA protein assay kit.
6
ACCEPTED MANUSCRIPT
2.7. Cell viability and cytotoxicity assays In the WST-1 assays, the HeLa cells (1.0 x 103) were seeded in a 96-well plate
PT
(Thermo Fisher Scientific) prior to the experiments for 24 hr. The medium of each well was replaced with DMEM containing 10% PB, 10% (v/v) fetal bovine serum, 1% (v/v)
SC RI
penicillin-streptomycin and 10 µM of each protein. After incubation for 4, 8, 24, or 48 hr, the medium of each well was replaced with the mixture of 100 µl of DMEM and 10 µl of the cell proliferation reagent WST-1 (Hoffmann-La Roche, Basel, Switzerland), and the cells were incubated for 2 hr. After incubation, the absorbance of each sample
NU
was measured using a Safire microplate reader (Tecan Group Ltd., Männedorf, Switzerland).
MA
In the LDH assays, the cells were incubated as well as the WST-1 assays for 24 hr and the detection of LDH activity was performed according to the manufacturer’s protocol using the CytoTox-ONE homogeneous membrane integrity assay (Promega).
ED
The cytotoxicity of each sample was calculated from the fluorescent intensities of the difference between the samples treated with reagents and the samples without reagents
PT
per the samples from lysed cells with reagents. 2.8. Statistical analysis
AC
CE
All statistical analysis was performed according to Welch’s t-test method.
3. Results
3.1. Construction and expression of EGFP-fusogenic peptide fusion proteins First, we designed and constructed EGFP fusion proteins with FPs derived from the sea urchin S. purpuratus mature bindin protein (B18, 103-120 aa and B55, 83-137 aa) (Fig. 1A). All proteins were expressed in E. coli and purified using Ni2+ affinity chromatography with >90% homogeneity as confirmed by the CBB staining of the SDS-PAGE gel (Fig. 1B). Next, we investigated the localization of purified EGFP-B18 and EGFP-B55 added to the HeLa cells. After a 4-hr incubation of each purified proteins with the HeLa cells, EGFP-B18 and EGFP-B55 were localized to the plasma membranes of the HeLa cells (Fig. 1C). Notably, EGFP-B55 was more strongly bound to the plasma membranes 7
ACCEPTED MANUSCRIPT
than EGFP-B18. This result is consistent with a previous report that the regions 69-91 aa and 119-130 aa of mature bindin were associated with the binding affinity of the
PT
peptide to the lipid bilayer [22]: the B18 peptide does not contain these regions, whereas the B55 peptide includes more than half of these regions, suggesting that the extended
SC RI
peptides of both sides of the B18 peptide contribute to its relatively higher affinity to lipid bilayers.
The low efficiency of the binding of the B18 peptide to the plasma membranes is common to other FPs, and FP-mediated endosomal escape has been confirmed by the
NU
additional fusion of CPPs to increase binding to plasma membranes [16,17]. Thus, we confirmed the function of EGFP-B18 using an additional fusion of a TAT peptide
MA
(Supplementary Fig. S1A). Although EGFP-TAT was primarily localized to endosome-like vesicles, most of the EGFP-B18-TAT was localized to the cytosol (Supplementary Fig. S1B). Quantitative image analysis revealed that the fluorescent
ED
intensity of intracellular EGFP-B18-TAT was approximately ten times higher than that of EGFP-TAT (Supplementary Fig. S1C). These results suggest that the B18 peptide
PT
facilitates the endosomal escape of EGFP-B18-TAT itself. Although the conjugation of a CPP-FP fusion peptide with a large protein, such as mCherry [27] or an antibody [28], sometimes reduces the endosomolytic activity of the CPP-FP peptide, the EGFP-fused
CE
B18-TAT peptide retained sufficient activity in facilitating endosomal escape. Therefore, the B18-TAT peptide is possibly suitable for the intracellular delivery of proteins
AC
directly conjugated to it.
8
AC
CE
PT
ED
MA
NU
SC RI
PT
ACCEPTED MANUSCRIPT
Fig. 1. Construction and expression of EGFP-FP fusion proteins. (A) Amino acid sequences of EGFP-fused fusogenic peptides (FPs). TCS; thrombin cleavage site. ‘GS’ dipeptide between EGFP and each additional peptide corresponds to a BamHI site. In EGFP-B18, there is an additional linker ‘PYD’ between EGFP and the B18 peptide (red), which is derived from parental bindin. (B) SDS-PAGE analysis of EGFP fusion proteins. EGFP fusion proteins were expressed in E. coli BL21(DE3) and purified using the Ni2+ beads column. The bands were detected using CBB staining. The molecular weights of EGFP, EGFP-B18, EGFP-B55 and EGFP-HA2 are 29.5, 32.1, 35.7 and 32.2 kDa, respectively. (C) Confocal microscopic images of the HeLa cells co-treated with 10 µM of the respective EGFP fusion protein (green) and 1 µg/ml Hoechst 33342 (blue) for 4 hr.
9
ACCEPTED MANUSCRIPT
3.2. EGFP-B55 promoted the intracellular delivery of dextrans Recently, Salomone et al. reported that a hybrid peptide of two antimicrobial
PT
peptides and a CPP from the HIV-1 Tat protein, CM18-TAT, had membrane-disruptive properties and enhanced the intracellular delivery of dextrans of various sizes by
SC RI
promoting endosomal escape [29]. Therefore, we investigated whether EGFP-B18-TAT and EGFP-B55 could enhance the plasma membrane permeability of dextrans. When EGFP-B55 and Texas Red-labeled dextrans were co-incubated with the HeLa cells, the dextrans were localized not only to endosome-like vesicles but also to both the
NU
cytoplasm and nucleus (Fig. 2A). The fluorescent intensity of the dextrans was significantly higher after co-incubation with EGFP-B55 than with EGFP (Fig. 2B).
MA
Histograms of each sample showed monodispersed peaks, indicating that the intracellular delivery of dextrans was promoted by EGFP-B55 in all incubated cells (Fig. 2C). This localization of EGFP-B55 and dextrans in fixed cells was similarly observed
ED
in live cells (Supplementary Fig. S2). In addition, the time-course experiments of the localization of EGFP-B55 and dextrans revealed that the amount of binding of
PT
EGFP-B55 to the cell membranes was gradually increased from 4 hr through 24 hr, while the intracellular delivery of dextrans in the presence of EGFP-B55 was rapidly increased between 8 and 24 hr (Supplementary Fig. S3). These results suggested that
CE
EGFP-B55 significantly promoted the intracellular delivery and endosomal escape of dextrans but that EGFP-HA2, a conventional FP derived from influenza virus
AC
hemagglutinin, did not affect dextran internalization. Because the activity of EGFP-B18-TAT was markedly less than that of EGFP-B55 (Supplementary Fig. S1D), the following analyses were performed on EGFP-B55.
10
AC
CE
PT
ED
MA
NU
SC RI
PT
ACCEPTED MANUSCRIPT
Fig. 2. Intracellular delivery of dextran 3000 in the presence of EGFP-FP fusion proteins. (A) Confocal microscopic images of the HeLa cells co-treated with 10 µM of the respective EGFP fusion protein (green), 60 µg/ml Texas Red-conjugated dextran 3000 (red) and 1 µg/ml Hoechst 33342 (blue) for 24 hr. Every medium had a pH of 7.1. The right panels correspond to the inset in left panels. (B) Relative fluorescent intensities of Texas Red per cell in the presence of each EGFP fusion protein. The intensities were normalized with respect to the average of the samples treated with EGFP. N = 59-75. Mean ± SD. * p
