Peptide-Induced DNA Condensation into Virus-Mimicking

Subscriber access provided by TUFTS UNIV

Biological and Medical Applications of Materials and Interfaces

Peptide-Induced DNA Condensation into Virus-Mimicking Nanostructures Meiwen Cao, Yu Wang, Wenjing Zhao, Ruilian Qi, Yuchun Han, Rongliang Wu, Yilin Wang, and Hai Xu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b00246 • Publication Date (Web): 06 Jul 2018 Downloaded from http://pubs.acs.org on July 7, 2018

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 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 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.

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 42 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

Peptide-Induced DNA Condensation into Virus-Mimicking Nanostructures Meiwen Cao,*,a Yu Wang,a Wenjing Zhao,a Ruilian Qi,b Yuchun Han,b Rongliang Wu,*,c Yilin Wang,b Hai Xu*,a

a

State Key Laboratory of Heavy Oil Processing and Centre for Bioengineering and

Biotechnology, College of Chemical Engineering, China University of Petroleum (East China), 66 Changjiang West Road, Qingdao 266580, China b

Key Laboratory of Colloid and Interface Science, Beijing National Laboratory for

Molecular Sciences (BNLMS), Institute of Chemistry, Chinese Academy of Sciences , Beijing 100190, China c

State Key Laboratory for Modification of Chemical Fibers and Polymer Materials,

College of Material Science and Engineering, Donghua University, Shanghai, China

Corresponding Authors * E-mail: [email protected] (M.C.); [email protected] (R.W.); [email protected] (H.X.)

1

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

Page 2 of 42

ABSTRACT A series of surfactant-like peptides have been designed for inducing DNA condensation, which are all comprised of the same set of amino acids in various different sequences. Results from experiments and molecular dynamics simulations show that the peptide’s self-assembly and DNA-interaction behaviors can be well manipulated through sequence variation. With optimized pairing modes between the β-sheets, the peptide of I3V3A3G3K3 can induce efficient DNA condensation into virus-mimicking structures. The condensation involves two steps, the peptide molecules firstly bind onto the DNA chain through electrostatic interactions and then self-associate into β-sheets under hydrophobic interactions and hydrogen bonding. In such condensates, the peptide β-sheets act as scaffolds to assist the ordered arrangement of DNA, mimicking the very nature of the virus capsid in helping DNA packaging. Such a hierarchy affords an extremely stable structure to attain the highly condensed state and protect DNA against enzymatic degradation. Moreover, the condensate size can be well tuned by the DNA length. The condensates with smaller sizes and narrow size distribution can deliver DNA efficiently into cells. The study helps not only for probing into the DNA packaging mechanism in virus but also delineating the role of peptide self-assembly in DNA condensation, which may lead to development of the peptide-based gene vectors for therapeutic applications.

KEYWORDS:

amphiphilic

peptide;

DNA

condensation;

virus-mimicking; gene delivery; molecular simulation

2


self-assembly;

Page 3 of 42 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


1. INTRODUCTION DNA is an important biomacromolecule that is negatively charged. The highly charged nature of DNA means that it readily adopts an extended conformation in solution due to electrostatic repulsion. However, it is desirable for DNA to package into a well condensed state in some specific cases.1,2 For example, DNA takes an ordered folding structure in the cellular nucleus (e.g. chromatin and ribosome) to allow for concentrated DNA storage, avoiding enzymatic digestion and performing complex biological processes.3-6 DNA compaction into a dense condensate is also required in nonviral gene therapy for facilitating the cellular uptake, protecting from nuclease degradation as well as overcoming endosomal entrapment so as to improve transfection efficiency.7-10

DNA condensation is usually mediated by certain condensing agents under cooperative non-covalent interactions and the choice of condensing agents can significantly impact the condensing mechanism and the resultant condensed nanostructure.11,12 For example, cationic surfactants/lipids can induce DNA compaction into globules or closely packed bead-like structures by first binding electrostatically onto the DNA chain and then causing DNA collapse into the compact state via hydrophobic interactions.13-19 Positively charged polyelectrolytes can induce DNA compaction into sphere-like aggregates through electrostatic interaction to generate polyion pairs and liberation of small counterions for entropy gain.9,11,20 The surfactant- or polyelectrolyte-induced DNA condensation usually results in random folding of the DNA chain. In contrary, nature gives the most efficient way for DNA condensation in a well-controlled manner to form highly ordered conformations. In the case of eukaryotes, DNA condensation involves a multistep hierarchical process 3



under which histone proteins package DNA into a precisely controlled beads-on-a-string structure.3-5 In the case of viruses, bacteria, and prokaryotes, DNA condensation is realized by interaction with the cationic proteins of spermine, spermidine and protamine, where the DNA chain takes ordered arrangements to form toroidal structures or bundles with regular hexagonal structure.21-25 These DNA condensates exhibit specific functionality by having a highly ordered molecular arrangement. One significant example is the superior ability of viruses as the most effective gene vectors.8 However, viruses have inherent toxic and immunogenic problems that are mainly due to the capsid proteins, which limits their clinical use.26 Therefore, it is a great desire to develop nonviral vectors that can micmic the typical functions of viruses whilst avoiding their drawbacks.9

Great efforts have been made in recent years to construct supramolecular assemblies that can mimic both the virus structures and functions, that is, artificial viruses.8,9,27 Short peptides have demonstrated several features which suggest they are excellent candidates for co-assembly with DNA to produce such artificial viruses.9, 28-31 As the peptides are made of natural amino acids, they are inherently biocompatible and non-toxic which makes them suitable for therapeutic use. Moreover, with modern synthesis techniques, short peptides can be easily designed and cheaply synthesized to produce sequences with specific functionality.32,33 This allows the design of self-assembling peptides which form structures that can mimic the folded protein conformations and functions for interacting with DNA to produce virus-like supramolecular assemblies.34,35 Recently, cocoon-like viral mimics were developed by Ni et al. through the co-assembly of a short designed peptide K3C6SPD with plasmid DNA. The morphology, stability, and function of the peptide/DNA co-assembled 4


Page 4 of 42

Page 5 of 42 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


structures were then tuned by regulating the inter-nanofibril hydrophobic interactions.31,36 Filamentous virus-like particles have also been produced by Ruff et al.

through

encapsulation

of

a

double-stranded

DNA

with

reassembled

mushroom-shaped nanostructures having a positively charged domain.37 Noble et al. have generated virus-like assemblies with self-assembling peptide shells as structural and functional mimetics of viruses.38 Moreover, the Glu-KW peptide was designed by Lim et al. with several functional segments to self-assemble into β ribbons, which can then interact with siRNA or dsDNA to produce filamentous artificial viruses.39 These studies constructed virus mimetics by taking advantage of the peptide/DNA supramolecular assembly and proved the importance of elaborate molecular design in the process. However, currently it is still a great challenge to construct artificial viruses that can mimic the natural viruses accurately by having highly ordered structural regularity and stability. It is also of great significance to get a comprehensive understanding of how the peptide molecular structure affects its co-assembly with nucleic acids so as to establish the guiding rules for peptide molecular design.

Recently we designed six surfactant-like peptides with the same amino acid composition, G3A3V3I3K3, K3I3V3A3G3, I3V3A3G3K3, K3G3A3V3I3, V3G3I3A3K3, and K3A3I3G3V3, as shown in Figure 1 and Figure S1 of Supporting Information.40 By adjusting the side chain group distribution on the two sides of a given β-sheet through sequence variation, the side-to-side pairing modes can be well manipulated to produce different self-assembled structures. The cone-like, dumbbell-like and irregular shaped peptides formed short nanorods, nanosheets and micrometer-long fibrils, respectively. These nanostructures exhibit different capacities in encapsulating insoluble 5



hydrophobic drug molecules and delivering them into cells.40 Very interestingly, here we show that these peptides can also induce efficient DNA condensation into highly condensed states and protect them from the enzymatic degradation. Specifically, I3V3A3G3K3 can interact with DNA to form virus-mimicking supramolecular assemblies by having the inter-sheet packing modes with high stability and suitable periodicity. In these co-assemblies, the peptide nano-sheets mimic viral capsids and the DNA chain wraps around them to package into well-ordered conformations. By having fine control over the molecular structures and assembly driving forces, these peptides provide an excellent system for probing into the peptide/DNA interactions. They also enable manipulation of the condensate structures by optimizing the peptide inter-sheet pairing modes so as to obtain the virus-mimicking peptide/DNA co-assembled structures.

Figure 1. (a) The amino acids used to construct the surfactant-like peptides. (b) The molecular structures of the designed peptides. The amino acids of G, A, V, and I that have the side alkyl chains of different size and hydrophobicity are used to construct 6


Page 6 of 42

Page 7 of 42 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


the hydrophobic part. The amino acid of K that has a side chain primary amine group is used to construct the hydrophilic part and provide positive charges. All peptides were N-terminal acetylated and C-terminal amidated. By simply adjusting the position of different segment units, three categories of molecules are produced, that is, the cone-like peptides of G3A3V3I3K3 and K3I3V3A3G3, the dumbbell-like peptides of I3V3A3G3K3 and K3G3A3V3I3, and the irregular shaped peptides of V3G3I3A3K3 and K3A3I3G3V3, respectively. The two peptides in one category are different in both the side chain group positions and the torsion angles by having reversed primary sequence, as seen from the stereochemistry shown in Figure S1.

1.

EXPERIMENTAL SECTION

2.1. Materials All of the peptides with purity > 96% were obtained from GL Biochem (Shanghai) Ltd. λ-DNA (M = 3.15×107, ca. 48.5 kbp) was obtained from New England BioLabs (Beijing, China). DNase of HindIII was purchased from Sigma-Aldrich. The green fluorescent protein (GFP) reporter gene of pEGFP-N2 was purchased from Biovector Co., Ltd (NTCC, Beijing). PCR primers for synthesis of a series of DNA with different lengths were designed by Sangon Biotech (Shanghai) Co., Ltd. Premix Taq enzyme was purchased from Takara. The DNA gel extraction kits were purchased from Tiangen Biochem Tech Co., Ltd. Biotin-labeled dUTP (Fluorescein-12-dUTP) was from Thermo Scientific. The other chemicals were purchased from Sinopharm Chemical Reagent Co., Ltd. All of the chemicals were used as received unless otherwise stated. All of the solutions were prepared with water (18.2 MΩ•cm) from a Milli-Q Biocel ultrapure water system.

7



2.2. Synthesis of DNA with different lengths A series of DNA with different lengths, that is, 300 bp, 900 bp, 1500 bp, and 2000 bp, were produced by PCR amplifying from the primers designed with plasmid pEGFP-N2 as template (Table 1). PCR amplifying was performed on a BIO-RAD T100TM Thermal Cycler. In parallel experiments, fluorescently labeled DNA of the above sequences were produced by using biotin-labeled dUTP (Fluorescein-12-dUTP) in the PCR amplifying process.

Table 1. Primers used for PCR production of DNA with different lengths. Product length

Sense primer

Antisense primer

300 bp

5’ CTCGTGACCACCCTGACCTA 3’

5’ GTTCTTCTGCTTGTCGGCCA 3’

900 bp

5’ GCTACCCCGACCACATGAAG 3’

5’ GCCGATTTCGGCCTATTGGT 3’

1500 bp

5’ GGCAGTACATCAATGGGCGT 3’

5’ GTTCACGTAGTGGGCCATCG 3’

2000 bp

5’ GCGATCACATGGTCCTGCTG 3’

5’ CGGCCACAGTCGATGAATCC 3’

2.3. Ethidium bromide (EtBr) displacement assay For EtBr displacement assay, 50 µL of λ-DNA (10 µg/mL) solution was first injected into each well of a black 96-well plate and then 50 µL of EtBr (2.5 µM) solution was added for mixing. After gentle shaking the mixed solutions were incubated in the dark for 10 min. Then 50 µL of peptide solution was injected into each well for incubation of 30 min. Next, the fluorescence emission at 600 nm was measured on a microplate reader (Molecular Instrument, M2e) with excitation at 520 nm. The fluorescence intensity of each sample was expressed as the ratio to the intensity from the DNA/EtBr mixed solution without peptide. Results averaged from four parallel 8


Page 8 of 42

Page 9 of 42 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


samples were calculated.

2.4. Dynamic light scattering (DLS) and zeta potential (ζ) measurements The hydrodynamic diameter and zeta potential measurements of λ-DNA and the λ-DNA/peptide complexes were measured on a Nano-ZS instrument (ZEN3600, Malvern Instruments, Worcestershire, UK) at room temperature. A clear disposable capillary cell (DTS1060C) was used for all of the experiments. Three repeats were performed for each sample.

2.5. Transmission electron microscopy (TEM) TEM characterization was carried out on a JEOL JEM-1400 equipment at 200 kV. The samples were prepared using negative staining and 10 µL of the sample solution was first deposited onto a parafilm surface. Then a copper grid with carbon Formvar-coating was placed on the solution drop for adsorption of approximately 8 minutes. After wiping away of the excess fluid with a filter paper the sample was placed on a drop of uranyl acetate solution (2% w/v) for staining which lasted approximately 5 minutes.

2.6. Atomic force microscopy (AFM) AFM characterization was performed on a Multimode Nanoscope VIII AFM (Bruker, Germany) under ambient environments. Scanasyst Air mode was used for image capturing. For sample preparation, 5–10 µL of sample solution was first deposited on a freshly cleaved mica surface and adsorbed for 10–60 sec. Then the sample was dried under a gentle nitrogen flow. Topographic images were recorded at 512 × 512 pixel. Image processing was performed with the vendor-supplied NanoScope Analysis 9



software.

2.7. Circular dichroism (CD) CD spectra were recorded on a MOS-450/AF-CD spectrophotometer (BioLogic, France) at room temperature using a 1.0-cm quartz cell. Scans were obtained in the range of 210‒340 nm at 1.0 nm intervals with an integration time of 0.5 s. Three repeats were performed for one sample and the averaged spectra were used.

2.8. Agarose gel electrophoresis Agarose gel electrophoresis was used to verify both retardation of the DNA immigration into the gel and protection of the DNA from nuclease digestion upon DNA/peptide complexing. For gel retardation experiments, the peptide/λ-DNA (10 µg/mL) mixed solutions at varied charge ratio R+/− (defined as the ratio of the peptide positive charges to the DNA negative charges) were tested. DNA Marker used is the 1 kb DNA ladder (Dye Plus) from TaKaRa Biotech., which is comprised of 10 dsDNA chains from 1 kbp to 10 kbp. For the nuclease digestion experiments, the DNase of HindIII was first added to the peptide/λ-DNA mixed solutions at varied R+/− for an incubation of 60 min at 37 °C. Then, the solutions were incubated at a higher temperature of 80 °C for another 20 min. For gel electrophoresis, 10 µL of the complexes were mixed with 2 µL of loading buffer and then loaded into the sample hole. Electrophoresis was run on a 1% (w/v) agarose gel containing EtBr (0.5 µg/ml) with TAE buffer at 110 V and the DNA location in the gel was photographed on a UV illuminator equipped with a Tanon 1600 camera.

2.9. Isothermal Titration Microcalorimetry (ITC) 10


Page 10 of 42

Page 11 of 42 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


The calorimetric measurements were carried out using a TAM III microcalorimetric system at 25.00 ± 0.01 °C with a stainless-steel sample cell of 1 mL. The sample cell was initially loaded with 600 µL of λ-DNA solution (50 µg/mL), and the peptide solution (1.0 mM) was consecutively injected into it at 10 µL per injection by a 500 µL Hamilton syringe under control of a 612 Thermometric Lund pump. The control calorimetric experiment was performed by titration of the peptide solution into water. Another cell loaded with 765 µL water was used as reference. The interval between two injections was 10 min for equilibrium. The solution was stirred with a gold propeller at 90 rpm and the experimental reproducibility was within ± 4%. The observed enthalpy change (∆Hobs) was calculated by integration of each injection peak in the P‒t (heat flow vs time) plot and corrected by subtracting the dilution enthalpy change for the titration of peptide solution into water.

2.10. Molecular dynamics (MD) simulations For the six peptides, one out of each category (cone-like, dumbbell-like or irregular) was chosen for performing MD simulations, since both peptides in the same category have similar self-assembly behaviors.40 The atomistic structures of I3V3A3G3K3, G3A3V3I3K3 and V3G3I3A3K3 were generated through Discovery Studio 3.5 in the β-strand conformation with the N-terminal acetylated and the C-terminal amidated. Figure 2a shows I3V3A3G3K3 as a representative. The molecule was positioned with the hydrogen bond direction aligned to the z-axis, and 10 such molecules were then replicated and rotated by 180 around the y-axis to form the antiparallel β-sheet structure (as reported in our recent work40) with initial inter-sheet distance of 0.5 nm (Figure 2b). The odd number of amino acids makes the upper (A) and lower (B) sides of this β-sheet have different hydrophobicity, thus the packing energy and stability of 11



each pairing conformation are of great importance for determining the self-assembly capacity as well as the self-assembled structures. Two such β-sheets were then placed at the initial inter-sheet distance of 1.2 nm with AA, AB and BB packing conformations, respectively. For example, Figure 2c shows the AA packing of the dumbbell-like molecule. The packed β-sheets were placed in a periodic box of 8.5 × 8.5 × 5 nm, where the strands at both ends could form hydrogen bonds with periodic images, resulting in an infinite fibril model of antiparallel β-sheets. Water molecules represented using the TIP3P water model were added to solvate the whole structure and 20 counter ions were added through randomly placing water molecules to neutralize the whole system. The CHARMM36 force field41 was used to represent the proteins and ions, and thousands of steps of energy minimization was performed before 2 ns MD simulations with heavy atoms restrained at their initial positions to equilibrate the water molecules. Then the side chains were further equilibrated for another 2 ns with the backbone heavy atoms restrained. Finally, 10 ns MD simulations were performed with no restraints, the temperature was coupled to 300 K using the velocity rescaling method and the pressures were semi-isotropically coupled to atmospheric pressure using the Berendsen method. A twin-range cutoff scheme was implemented between 1.2-1.4 nm for the non-bonded interactions and the particle mesh Ewald method42 with a fourier spacing of 0.1 nm was applied for the long range electrostatic interactions. All covalent bonds with hydrogen atoms were constrained using the LINCS algorithm.43 All MD calculations and system building were performed using the GROMACS44 MD simulation package version 5.0. Detailed parameters used for the modeling can be found in Tables S1-S4 of Supporting Information.

12


Page 12 of 42

Page 13 of 42 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 2. a) I3V3A3G3K3 molecules in the β-strand conformation; b) the antiparallel β-sheet structure generated from 10 molecules positioned at the inter-strand distance of 0.5 nm; c) the AA surface packing of two β-sheets at the inter-sheet distance of 1.2 nm; d) the overall view and e) the cross-sectional view of the infinite fibril model used in MD simulation with hydrogen bonds with periodic images in the Z-axis. The pink molecules represent the water molecules and the green spheres represent the counter-ions.

2.11. Cytotoxicity evaluation The cytotoxicity of the peptide/DNA complexes was evaluated by the tetrazolium reduction (MTT) assay. Firstly 100 µL of cells were added to each well of a 96-well plate at a density of 1 × 105 cells/mL. The cells were then cultured at 37 °C for 24 h to become adherent. After that the complete medium was removed from each well and renewed with 100 µL of new medium. Then 100 µL of the peptide/DNA mixed solution at varied R+/− was added into each well. The cells were cultured for another 24 h and 20 µL of MTT solution (5mg/mL) was added to each well. After further 13



incubating at 37 ºC for 4 h, the supernatant was thoroughly aspirated and mixed with 150 µL of DMSO by shaking for 10 min. The absorbance at 570 nm was collected on a microplate reader (Molecular Instrument, M2e) and used to calculate the cell survival ratio (P) with the following formula:

= 1 − − / / − 1

where ATris is absorbance of the medium, Apeptide/DNA is absorbance of the mixed solution of the peptide/DNA complexes and the medium, and Ablank is absorbance of DMSO. The cells cultured in Tris buffer were used as control. Results averaged from four parallel samples were calculated.

2.12. Gene delivery and transfection A series of DNA with different lengths, that is, 300 bp, 900 bp, 1500 bp, and 2000 bp, were used to form complexes with I3V3A3G3K3 for measurements of both the morphology and the gene delivery efficiency. 293E cells were cultured in the 96-well plate for 24 h. The I3V3A3G3K3/DNA mixtures at R+/− of 10, 20, and 50 were added into each well and stabled for 20 min. Then 50 µL of mixed solution and 100 µL of serum-free DMEM were added per well. After 6 h serum was added to each well and incubation for another 24 to 48 h. The delivery efficiency was evaluated either by imaging on an inverted fluorescence microscope (Leica DMi8) or analysis on a FACSCalibur flow cytometer (BD Biosciences, USA). Non-treated cells were used as negative controls.

14


Page 14 of 42

Page 15 of 42 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


The peptide-mediated gene transfection was tested by using the reporter gene of pEGFP-N2, a GFP reporter gene. 293E cells were cultured in a 24-well plate to become adherent. 3.0 µg of peptide and 0.8 µg of pEGFP-N2 were separately dissolved in 50 µl of DMEM and incubated at room temperature for 5 min. The peptide and pEGFP-N2 solutions were then mixed and incubated for 20 min. After removing the cell culturing medium from each well and renewed with 500 µl of fresh serum-free medium, 100 µl of the peptide/pEGFP-N2 mixed solution was injected into each well. After incubation for another 4 h, the medium was renewed with another 500 µl of fresh fetal bovine serum-containing medium and for incubation of another 24 h. After that, the cells were imaged on an inverted fluorescence microscope (Leica DMi8) to check the GFP expression. Polyethyleneimine (PEI), a classical cationic polymer and an accepted standard for gene transfection, was used as control in a 3:1 ratio of PEI to pEGFP-N2 (w/w).

3. RESULTS AND DISCUSSION 3.1. Peptide-mediated DNA condensation The EtBr displacement assay was first used to assess the binding affinity of different peptides to DNA. EtBr can intercalate into DNA and give greatly enhanced fluorescence emission at 610 nm. However, when the condensing agents bind to DNA they can displace the intercalated EtBr and result in reduction of the fluorescence intensity.45-47 Figure 3a shows that five peptides lead to significant decreases of the fluorescence intensity with the increase of R+/−, showing strong DNA binding affinity, the exception being K3A3I3G3V3. By comparing the relative fluorescence intensity at equilibrium, the binding affinity of different peptides to DNA is ordered as I3V3A3G3K3 > G3A3V3I3K3 > K3I3V3A3G3 ≈ K3G3A3V3I3 > V3G3I3A3K3 > 15



Page 16 of 42

K3A3I3G3V3 (Table 1). Specifically, I3V3A3G3K3 gave an extremely low relative fluorescence intensity value of ~0.18. Moreover, it also had the smallest R+/− at the transition point, which was ~2.8. The results demonstrate well that I3V3A3G3K3 has superior DNA binding and condensing ability.

Figure 3. (a) The EtBr displacement assay showing the binding of peptides with DNA. (b) Variation of ζ and hydrodiameter of the I3V3A3G3K3/DNA complexes as a function of R+/−.

Table 1. Relative fluorescence intensity at equilibrium (RFLIE) and R+/− value at the transition point to fully displacement (R+/−(T)) of EtBr, as derived from Figure 1a. peptides

G3A3V3I3K3

K3I3V3A3G3

I3V3A3G3K3

K3G3A3V3I3

V3G3I3A3K3

K3A3I3G3V3

RFLIE

0.39±0.03

0.49±0.02

0.18±0.01

0.49±0.05

0.60±0.10

0.90±0.10

R+/−(T)

~3.5

~3.3

~2.8

~3.7

~3.4

With I3V3A3G3K3 as a typical representative, we investigated the variation of ζ and hydrodiameter of the peptide/DNA complexes with the increase of R+/− (Figure 3b). With R+/− < 1, ζ kept the negative values of about ‒30 mV and the hydrodiameter remained large at 820 ± 60 nm. With 1 < R+/− < 3, ζ increased with increasing R+/−, and the negative-to-positive transition occurred roughly at R+/− of 2.5. In this region, 16


Page 17 of 42 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


the hydrodiameter decreased gradually and got to 210 ± 50 nm at R+/− of 3.0. Then, with R+/− >3, ζ reached equilibrium of about +30 mV and the hydrodiameter decreased to 100‒220 nm. The results clearly show that the cationic peptide molecules bound successfully on the negatively charged DNA and caused charge reversal. Such an interaction leads to efficient DNA condensation into highly compacted conformations.

TEM was further used to investigate the morphologies of λ-DNA and the peptide/λ-DNA complexes (Figure 4). Figure 4a shows that λ-DNA took an extended conformation in solution. However, some aggregates with irregular shapes appeared after addition of the peptide into the solution, as indicated by the yellow arrows in Figure 4b-e. Since the peptide concentration of 10 µM is greatly lower than the critical aggregation concentration (CAC) of 590, 650, 530, and 280 µM for G3A3V3I3K3, K3I3V3A3G3, I3V3A3G3K3, and K3G3A3V3I3, respectively,40 these aggregates are ascribed to the peptide/λ-DNA co-assembled structures (DNA condensates) rather than the peptide self-assembled structures. Interestingly, some ordered domains can be found, as shown by the redrawn cartoon images in Figure 4b, d, and e. The peptide/λ-DNA complexes took different morphologies, sizes and ordered domain levels in each case (Table 2 and Figure S2 of Supporting Information), which may affect their delivery efficiency. In the cases of K3I3V3A3G3 and K3G3A3V3I3, the complexes gave loose aggregates with relatively larger sizes and lower levels of ordered domains, while in the cases of G3A3V3I3K3 and I3V3A3G3K3 the complexes gave dense aggregates with smaller sizes and higher levels of ordered domains. Specifically, the I3V3A3G3K3/λ-DNA condensates showed the smallest sizes of 150 ± 75 nm and the highest content of the ordered domains. In fact, the ordered 17



domains dominated the condensates morphology, as shown in Figure 4e. The results confirm the highest DNA condensing efficiency of I3V3A3G3K3, and the ordered domain formation suggests a specific co-assembly behavior and interaction mode with DNA, as will be discussed in the following text.

Figure 4. TEM images of (a) λ-DNA (10 µg/mL) and the peptide/λ-DNA complexes at R+/− of 3 (peptide: 10 µM, λ-DNA: 10 µg/mL) for (b) G3A3V3I3K3, (c) K3I3V3A3G3, (d) K3G3A3V3I3, and (e) I3V3A3G3K3, respectively. The insets of (b), (d), and (e) are the redrawn cartoons that display the orderly packed domains in the corresponding squared areas. For V3G3I3A3K3 and K3A3I3G3V3 the data are absent because no obvious co-assembled structures can be obtained.

18


Page 18 of 42

Page 19 of 42 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


Table 2. Morphologies, sizes and ordered domain levels of the peptide/λ-DNA condensates as obtained from the TEM results. complex

morphology

size (nm)

ordered domain level

G3A3V3I3K3/λ-DNA

dense particles

180 ± 100

medium

K3I3V3A3G3/λ-DNA

loose aggregates

300 ± 220

low

I3V3A3G3K3/λ-DNA

dense particles

150 ± 75

high

K3G3A3V3I3/λ-DNA

loose aggregates

250 ± 160

low

V3G3I3A3K3/λ-DNA

hardly found

‒

‒

K3A3I3G3V3/λ-DNA

not found

‒

‒

Further efforts were made to investigate the details of the I3V3A3G3K3-induced DNA condensates. Figure 5a showed many dense particles in a single image, indicating the popularity of the condensates and the high condensing efficiency. The AFM image (Figure 5h) showed three-dimensional morphologies of the condensates with varied morphologies and sizes, in good agreement with the TEM results. These structures were quite stable and retained similar morphologies in an incubation time range of one month. And the DSC profile showed that the complexes were stable at temperature below 60 °C (Figure S3, Supporting Information). Then, as can be seen from Figure 5b-f, these condensates might take quite different shapes, for example, nanospheres (b), wormlike structures (c), bridged structures (d), irregular-shaped structures (e), and squared shapes (f). Interestingly, these structures all had some ordered domains that acted as skeletons to assist the whole structure. The periodicity of the ordered domains was about 5.5 nm from the sectional analysis (Figure 5g). Moreover, the FTIR spectrum of the I3V3A3G3K3/λ-DNA complexes showed a peak at 1629 cm-1 and the CD spectrum showed a negative minimum at 218 nm (Figure 5i), all typical of the β-sheet secondary structures, indicating that the I3V3A3G3K3 19



molecules associated into ordered β-sheets. As reported in our recent study, the β-sheet length in the strand direction was about 4.0 nm (corresponding to the lamella thickness) for I3V3A3G3K340 and the diameter of λ-DNA is ~2.0 nm. This suggests the ordered β-sheets and DNA molecules are packing together to form DNA-wrapped β-sheets in the DNA condensates, with periodicity ~5.5 nm. In such DNA condensates, the peptide β-sheets act as scaffolds to assist the ordered packaging of the DNA chain. As known, the viruses are characterized by the presence of protein shells (capsids) and condensed DNA (or RNA) in the core.48 The capsids consist of oligomeric structural subunits made of protomers, which arrange into highly ordered helical or icosahedral structure.49,50 The capsids further enclose DNA or RNA to form hierarchical supramolecular structures via non-covalent interactions. Here, the peptide/DNA condensates mimic well the structural nature of the viruses, where the β-sheets act as protomers to assemble into capsid-like higher-order structures and further assist the packaging of DNA to form supramolecular nanostructures. Similarly, Ni et al. have produced cocoon-like virus mimics with regular striped surface patterns by designing peptides for co-assembly with DNA.31,36 These results all estabilish that the rationally designed peptide/DNA supramolecular assemblies can form highly ordered structures to mimic well the virus structures.

20


Page 20 of 42

Page 21 of 42 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 5. TEM (a‒f) and AFM (h) images as well as the FTIR and CD spectra (i) of the I3V3A3G3K3/DNA complexes at R+/− of 3, (a) the overall TEM image showing several I3V3A3G3K3/DNA complexed nanostructures, as indicated by the red circles, (b‒f) the magnified images showing details of each structure with different shapes, and (g) the gray-scale section analysis corresponding to the red line in image f.

The ability of I3V3A3G3K3 to induce DNA condensation was further validated by the agarose gel retardation assay. Figure 6a shows the agarose-gel electrophoretic mobility of the I3V3A3G3K3/DNA complexes at varied R+/−. A single bright band was observed for λ-DNA (lane 1), and it became faint with increasing R+/− (lanes 2, 3, and 21



4) and nearly disappeared at R+/− of 2.4. Such a band moved slower than the DNA marker with the largest molecular mass (10 kbp) as verified in a control electrophoresis experiment with DNA markers (Figure S5, Supporting Information), indicating it was from λ-DNA with a larger molecular mass (48.5 kbp). The results showed successful retardation of the DNA migration by I3V3A3G3K3/DNA complex formation. Nearly complete retardation was achieved at R+/− of 2.4. Then, to assess whether the peptide/DNA complexation can protect DNA from the nuclease digestion, we incubated both bare DNA and the peptide/DNA complexes with an endonuclease of HindIII and the results were also analyzed by agarose gel electrophoresis.30 As shown in Figure 6b, the bare DNA produced several electrophoretic bands (lane I), indicating that λ-DNA was degraded into several short segments. However, for the peptide/DNA complexes, the bands corresponding to degraded segments became faint gradually with increasing R+/− and no bands were found at R+/− > 2.4, indicating the successful inhibition of nuclease digestion. In parallel experiments, it was found that the condensates can also protect DNA from digestion by exonuclease III (Figure S5, Supporting Information). All these results confirm that the co-assembled structures afford an excellent stability to effectively protect DNA against enzymatic degradation.

Figure 6. Agarose gel electrophoresis results showing (a) the binding of I3V3A3G3K3 22


Page 22 of 42

Page 23 of 42 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


with DNA and (b) protection against nuclease digestion at pH 7.0, lane 1: λ-DNA, lanes 2, 3, and 4: the I3V3A3G3K3/λ-DNA complexes at R+/− of 1.0, 2.4, and 3.2, respectively; lane I: λ-DNA + HindIII (DNase), lanes II, III, and IV: the I3V3A3G3K3/λ-DNA complexes + HindIII, with R+/− of 1.0, 2.4, and 3.2, respectively.

Moreover, further gel electrophoresis with samples at different pH (Figure S5, Supporting Information) showed that, successful DNA retardation can be realized at pH of 5.0 and 7.0, while no obvious retardation can be found at pH 8.4. The results clearly disclosed the effect of pH on the peptide/DNA complex formation. As known, the peptide positive charges come from the primary amine groups (‒NH2) of lysine, and the protonation of ‒NH2 depends on pH, therefore, the peptide will have more positive charges at lower pH due to higher protonation of ‒NH2. It is reasonable that at pH 5.0/7.0 the peptide had sufficient positive charges for a strong electrostatic interaction with DNA, while at pH 8.4 the peptide had less positive charges and cannot afford a stable electrostatic binding with DNA.

ITC was employed to investigate the thermodynamic behavior of the interaction between I3V3A3G3K3 and DNA.18 Figure 7a gives the ITC curve of the observed enthalpy change (∆Hobs) against the peptide concentration for titration of peptide into the DNA solution (Figure S6, Supporting Information). The enthalpy profile consists of an initial endothermic region followed by a sharp rise in enthalpy, and then the enthalpy drops sharply to a nearly constant value. The endothermic heat in the rising region should be attributed to dehydration due to electrostatic binding of positively charged I3V3A3G3K3 with negatively charged DNA. The saturation point of electrostatic binding is indicated by the endothermic maximum. In the decreasing 23



region, the enthalpy values changed towards the less endothermic because of the peptide-induced DNA aggregation, mainly aided by hydrophobic interaction and hydrogen bonding. The stability of the I3V3A3G3K3/DNA complexes can also be inferred from the binding constant (Ka), which was calculated to be about 6.34 × 106 M−1 from the ITC result. The value is an order of magnitude higher than the reported Ka of the PEI@plasmid DNA complexes,51 indicating that I3V3A3G3K3 has a stronger interaction with DNA than PEI, a classical cationic polymer for DNA binding and delivery.51

Figure 7. (a) The ITC curve of observed enthalpy change (∆Hobs) against the final peptide concentration as obtained from the titration of I3V3A3G3K3 into λ-DNA solution at 25 °C. (b) The proposed DNA condensing pathway induced by the designed peptides. The peptide associates to the DNA and this forms regions along the DNA where the peptide concentration is high enough to form β-sheets, stabilizing the aggregates and leading to the final virus-like structures.

3.2. Condensing mechanism 24


Page 24 of 42

Page 25 of 42 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


Considering that the effective peptide concentration (

Peptide-Induced DNA Condensation into Virus-Mimicking

Recommend Documents