Encapsulation of mRNA into Artiﬁcial Viral Capsids via Hybridization of a β -Annulus-dT 20 Conjugate and the Poly(A) Tail of mRNA

: Messenger RNA (mRNA) drugs have attracted considerable attention as promising tools with many therapeutic applications. The e ﬃ cient delivery of mRNA drugs using non-viral materials is currently being explored. We demonstrate a novel concept where mCherry mRNA bearing a poly(A) tail is encapsulated into capsids co-assembled from viral β -annulus peptides bearing a 20-mer oligothymine (dT 20 ) at the N-terminus and unmodiﬁed peptides via hybridization of dT 20 and poly(A). Dynamic light scattering measurements and transmission electron microscopy images of the mRNA-encapsulated capsids show the formation of spherical assemblies of approximately 50 nm. The encapsulated mRNA shows remarkable ribonuclease resistance. Further, modiﬁcation by a cell-penetrating peptide (His16) on the capsid enables the intracellular expression of mCherry of encapsulated mRNA. of synthesis of dT 20 modified β -annulus peptide (dT 20 -SS- β -annulus) and formation of an artificial viral capsid by co-assembly of β -annulus peptide and dT 20 -SS- β -annulus hybridized with mRNA. 20 -SS- β -annulus hybridized with mRNA.


Introduction
Therapeutic mRNA have attracted attention in recent years as a new type of nucleic acid drugs. Such mRNAs show potential for use in protein replacement therapy and vaccination [1][2][3][4][5][6]. mRNA drugs have the advantage of inducing direct protein synthesis in the cytoplasm, thus differing from DNA therapy. Naked mRNA is; however, unstable outside of cells and is unable to effectively penetrate cell membranes due to electrostatic repulsion. Thus, various materials for mRNA delivery, such as liposomes, dendrimers, and polyion complex micelles, have been developed and are attractive for their flexibility in molecular design [2][3][4][5][6][7][8][9][10][11]. Conversely, RNA viruses package specific genome molecules inside their outer protein shell, the viral capsid [12,13]. Viral capsids are also attractive for precise mRNA packaging and delivery but the use of natural viruses for mRNA delivery has safety concerns. Virus-like artificial protein cages for RNA packaging are, therefore, attractive. Recently, Hilvert and coworkers demonstrated that a non-viral protein cage formed by Aquifex aeolicus lumazine synthase selectively packages mRNA via cationic peptide tags [14,15].
Eukaryotic mRNA possesses a long adenine nucleotide (poly(A)) tail of about 100~200 nt at the 3 -end to regulate translation [16][17][18][19][20][21]. We propose a novel strategy for the encapsulation of mRNA into virus-like peptide nanocapsules via hybridization between dT 20 and the poly(A) tails. We developed an "artificial viral capsid" with a size of 30-50 nm self-assembled from the β-annulus Figure 1. Schematic of synthesis of dT20 modified β-annulus peptide (dT20-SS-β-annulus) and formation of an artificial viral capsid by co-assembly of β-annulus peptide and dT20-SS-β-annulus hybridized with mRNA.

General
Reverse-phase HPLC was performed at ambient temperature with a Shimadzu LC-6AD liquid chromatograph equipped with a UV-Vis detector (220 nm and 260 nm, Shimadzu SPD-10AVvp, Kyoto, Japan) using an Inertsil WP300 C18 column (GL Science, 250 mm × 4.6 mm or 250 × 20 mm). MALDI-TOF mass spectra were obtained on an Autoflex T2 (Bruker Daltonics, Billerica, USA) in linear/positive mode with α-cyano-4-hydroxycinnamic acid (α-CHCA) or 3-hydroxypicolinic acid (3-HPA) with diammonium hydrogen citrate as the matrix. The UV-Vis spectra of DNA-conjugated peptides were measured at 260 nm using a Jasco V-630 (JASCO Corporation, Tokyo, Japan) with a quartz cell (S10-UV-1, GL Science). An mRNA coding mCherry fluorescent protein (mCherry mRNA, 1 kb) with a 3 -polyadenylic acid (poly(A)) tail was purchased from OZ Biosciences (Marseille, France). RNase A was purchased from Nacalai Tesque (Kyoto, Japan). All other reagents were obtained from a commercial source and used without further purification. Deionized water of high resistivity (>18 MΩ cm) was purified using a Millipore Purification System (Milli-Q water, Merck Millipore, Burlington, USA) and used as a solvent.
An amine-modified dT 20 (5 NH 2 -(CH 2 ) 6 -TTTTTTTTTTTTTTTTTTTT-3 , Gene Design Inc., Osaka, Japan) in 50 mM sodium phosphate buffer (pH 8.0) was added to 20-fold molar excess of N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) in acetonitrile and the mixture was incubated for 1 h at 25 • C. A Spectra/por7 membrane with a cutoff Mw 1000 (Spectrum Laboratories, Inc., Rancho Dominguez, USA) was used for dialysis against water for 24 h. The internal solution was lyophilized to afford a flocculent solid. This product was dissolved in water and the concentration was defined by UV-Vis spectroscopy. Product (PySS-dT 20  An aqueous solution of PySS-dT 20 (0.1 mM) in a 50 mM sodium phosphate buffer (pH 7.2) was mixed with the Cys-β-annulus peptide (2 mM) in a 50 mM sodium phosphate buffer (pH 7.2) and the mixture was incubated for 24 h at 25 • C. The solution was purified by reverse-phase HPLC eluted with a linear gradient of CH 3 CN/0.1 M ammonium formate aqueous solution (10/90 to 100/0 over 95 min). The elution fraction was collected, concentrated in a centrifugal evaporator, and dialyzed as above against water for 20 h. The internal solution was lyophilized to afford a flocculent solid. This product was dissolved in water and the concentration was defined by UV-Vis spectroscopy. Product (dT 20 -SS-β-annulus peptide) yield was 12.6 nmol (27.4%)-MALDI-TOF-MS (matrix: 3-HPA): m/z = 8710 (exact mass: 8698).
Dynamic light scattering (DLS) was measured with a Zetasizer NanoZS (Malvern, Kobe, Japan) instrument at 25 • C using an incident He-Ne laser (633 nm). Correlation times of scattered light intensities G(τ) were measured several times and the means were fitted to Equation (1), where B is the baseline, A is amplitude, q is the scattering vector, τ is the delay time, and D is the diffusion coefficient.
Hydrodynamic radii (R H ) of scattering particles were calculated using the Stokes-Einstein Equation (2), where η is the solvent viscosity, k B is Boltzmann's constant, and T denotes absolute temperature.
Transmission electron microscopy (TEM) images were obtained with a JEOL JEM 1400 Plus (JEOL Ltd., Tokyo, Japan), using an accelerating voltage of 80 kV. Aliquots (5 µL) of DLS samples were applied to hydrophilized carbon-coated Cu-grids (C-SMART Hydrophilic TEM girds, Alliance Biosystems, Osaka, Japan) for 60 s and then removed. Subsequently, the TEM grid was instilled in the aqueous solution (5 µL) of 2% phosphotungstic acid, Na 3 (PW 12 O 40 )(H 2 O) n . The staining solution was removed after 60 s and the sample-loaded grids were dried in vacuo.

Electrophoretic Mobility Shift Assay and Nuclease Resistance Assay
Electrophoretic mobility shift assay (EMSA) was employed for the detection of peptide complexes with nucleic acids. The dT 20 -SS-β-annulus:β-annulus/mRNA solutions were loaded onto 3% w/v agarose gels in a TAE buffer, pre-cast with GelRed (Wako Pure Chemical Industries, Osaka, Japan) for nucleic acid detection. Three microliters of the dT 20 -SS-β-annulus:β-annulus/mRNA solution ((mRNA nt) = (T) = 1 mM) was loaded with 5 µL of Bluejuice Gel loading buffer (Thermo Ficher, Waltham, USA). Electrophoresis used 210 V for 30 min on an Atto AE-6100 with Atto mypower II300 AE-8130 (Atto, Tokyo, Japan). The bands were visualized with a UV illuminator (TP-15 MP, Atto, Tokyo, Japan) and the images were recorded with a digital camera.
Ribonuclease A from bovine pancreas (Nacalai Tesque, Kyoto, Japan) is an endoribonuclease which specifically hydrolyzes the 3 end of pyrimidine residues in single-stranded RNA. Each peptide/mRNA solution was incubated with a solution of RNase A (3 µL, 1 U/µL) in PBS for 10-60 min at 37 • C. Enzymatic degradation of mRNA was evaluated by EMSA.

Confocal Laser Scanning Microscopy (CLSM) Measurements of In-Cell Expression of mCherry mRNA
Confocal laser scanning microscopy (CLSM) used a Fluo View FV10i (Olympus, Tokyo, Japan). Human hepatoma HepG2 cells were cultured in DMEM (10 v/v% FBS, 100 µg/mL streptomycin, 100 units/mL penicillin, 1 mM sodium pyruvate, and 1 v/v% MEM nonessential amino acids) for 24 h at 37 • C in a 5% CO 2 atmosphere. The cells were seeded onto single-well bottom dishes with 2.0 × 10 4 cells/well in a final volume of 100 µL and incubated for 24 h at 37 • C and 5% CO 2 . The medium was removed and 50 µM dT 20 -SS-β-annulus, 450 µM β-annulus, and 1 µM mCherry mRNA in the fresh medium were added to the cells, then incubated for 48 h under 5% CO 2 . The co-assembly mixture of 50 µM dT 20 -SS-β-annulus, 450 µM β-annulus-His 16 , and 1 µM mCherry mRNA was added to the cells for evaluation of expression of mCherry fluorescent protein. A complex of 2 µM TransIT-mRNA (Transfection kit, Mirus Bio LLC, Madison, USA) with 1 µM mCherry mRNA was added to cells as a positive control. CLSM images of mCherry fluorescence were obtained with excitation at 587 nm and an mCherry band-pass filter (Red). Fluorescence intensity was measured by cell from CLSM images by subtracting background intensity using ImageJ 1.51 software.

Synthesis and Self-Assembling Behavior of β-Annulus Peptide Bearing dT 20 at the N-Terminus
We designed a β-annulus peptide bearing dT 20 at the N-terminus that was directed to the interior of the capsid (Figure 1) to encapsulate mRNA via hybridization between dT 20 and the poly(A) tail at the 3 end of mRNA. A 5 -terminal aminated 20-mer thymidine nucleotide (dT 20 -NH 2 ) was reacted with N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) to obtain a dT 20 bearing pyridyl disulfide group at the 5 -end (PySS-dT 20 ). A Cys-β-annulus peptide (CINHVGGTGGAIMAPVAVTRQLVGS) containing Cys at the N-terminus was synthesized by standard Fmoc-based solid-phase methods according to our reported procedure [28]. PySS-dT 20 was attached to the Cys-β-annulus peptide with a disulfide-exchange reaction. A reversed-phase HPLC chart of the reaction mixture showed one peak at 26.7 min, which was different from the retention time of dT 20 -NH 2 and PySS-dT 20 (Figure 2A). The purified product was assigned by MALDI-TOF-MS as dT 20 -modified β-annulus peptide (dT 20 -SSβ-Annulus) ( Figure 2B). DLS of a 50 µM solution of dT 20 -SS-β-annulus in PBS (pH 7.4) exhibited formation of 77 ± 21 and 40 ± 8 nm assemblies ( Figure 2C). TEM images of the aqueous solution stained with phosphotungstic acid showed the formation of spherical assemblies of approximately 40 nm in diameter ( Figure 2D). The concentration dependence of the size distribution in DLS measurement indicated that dT 20 -SS-β-annulus formed assemblies with sizes ranging from 30 to 100 nm in a concentration range of 5-125 µM ( Figure S1). This tendency was similar to the findings in our previous report [35]. Notably, β-annulus peptide modified with highly charged dT 20 formed relatively stable artificial viral capsids.

Complexation of mCherry mRNA and Artificial Viral Capsid Bearing dT20
We initially attempted encapsulation of mRNA within the artificial viral capsid consisting of dT20-SS-β-annulus via hybridization between dT20 and mRNA. However, when poly(A) and dT20-SSβ-annulus were mixed at equimolar base concentrations, capsid formation was insufficient (data are not shown). It is probably difficult to self-assemble among dT20-SS-β-annulus peptides hybridized on a large poly(A) molecule due to the excluded volume effect. Therefore, we developed an alternate strategy to encapsulate mRNA with dT20-SS-β-annulus and unmodified β-annulus peptide (INHVGGTGGAIMAPVAVTRQLVGS) (Figure 1). DLS and TEM images of the mixture of dT20-SS-βannulus/unmodified β-annulus peptide at 1:9 molar ratio in PBS showed the formation of spherical co-assemblies of size 45 ± 20 nm (Figure 3). The peptides mixture at other molar ratios (dT20-SS-βannulus:β-annulus = 1:4.5, 1:2, 1:1) also formed spherical co-assemblies of similar size ( Figure S2). Next, we encapsulated mRNA coding in the mCherry fluorescent protein (mCherry mRNA) by coassembly at different molar ratios of dT20-modified and unmodified β-annulus peptide.

Complexation of mCherry mRNA and Artificial Viral Capsid Bearing dT 20
We initially attempted encapsulation of mRNA within the artificial viral capsid consisting of dT 20 -SS-β-annulus via hybridization between dT 20 and mRNA. However, when poly(A) and dT 20 -SS-β-annulus were mixed at equimolar base concentrations, capsid formation was insufficient (data are not shown). It is probably difficult to self-assemble among dT 20 -SS-β-annulus peptides hybridized on a large poly(A) molecule due to the excluded volume effect. Therefore, we developed an alternate strategy to encapsulate mRNA with dT 20 -SS-β-annulus and unmodified β-annulus peptide (INHVGGTGGAIMAPVAVTRQLVGS) (Figure 1). DLS and TEM images of the mixture of dT 20 -SS-β-annulus/unmodified β-annulus peptide at 1:9 molar ratio in PBS showed the formation of spherical co-assemblies of size 45 ± 20 nm (Figure 3). The peptides mixture at other molar ratios (dT 20 -SS-β-annulus:β-annulus = 1:4.5, 1:2, 1:1) also formed spherical co-assemblies of similar size ( Figure S2). Next, we encapsulated mRNA coding in the mCherry fluorescent protein (mCherry mRNA) by co-assembly at different molar ratios of dT 20 -modified and unmodified β-annulus peptide.
(INHVGGTGGAIMAPVAVTRQLVGS) (Figure 1). DLS and TEM images of the mixture of dT20-SS-β-annulus/unmodified β-annulus peptide at 1:9 molar ratio in PBS showed the formation of spherical co-assemblies of size 45 ± 20 nm (Figure 3). The peptides mixture at other molar ratios (dT20-SS-βannulus:β-annulus = 1:4.5, 1:2, 1:1) also formed spherical co-assemblies of similar size ( Figure S2). Next, we encapsulated mRNA coding in the mCherry fluorescent protein (mCherry mRNA) by coassembly at different molar ratios of dT20-modified and unmodified β-annulus peptide. An aqueous solution of mCherry mRNA in PBS was added to the lyophilized powder of dT 20 -SS-β-annulus:β-annulus peptide followed by dilution with PBS and incubation for 30 min at 25 • C to prepare capsids containing mCherry mRNA. The electrophoretic mobility shift assay (EMSA) showed a band of mCherry mRNA significantly shifted by complexation with a co-assembly mixture of dT 20 -SS-β-annulus:β-annulus peptide at a 1:9 molar ratio ( Figure 4A, lane 4). In contrast, the other co-assembly mixtures showed only a slight retardation of mRNA ( Figure 4A, lane 5-7), which reflect almost neutral net charge and relatively small size of the β-annulus peptide. The migration position of a mixed solution of unmodified β-annulus peptide and mCherry mRNA was not shifted from mCherry mRNA alone ( Figure 4B). DLS and TEM images of the 1:9 co-assembly complexes with mCherry mRNA show the formation of spherical assemblies of size 52 ± 10 nm ( Figure 5A,B). We previously reported that the N-terminus of the β-annulus peptide is directed toward the interior of the capsid and the surface ζ-potential of the capsid is almost zero at neutral pH [25]. Thus, when negatively charged RNA is encapsulated at neutral pH, the artificial viral capsid should minimally migrate from the applied position due to the capsid charge shield. Therefore, EMSA results indicate that the dT 20 -SS-β-annulus:β-annulus peptide 1:9 co-assembly encapsulated mCherry mRNA via hybridization with dT 20 directed to the interior of capsids. EMSA shows that mCherry mRNA could be complexed with the 1:9 co-assembly at 4 • C, 25 • C, 37 • C, and 60 • C ( Figure S3). However, complexation at 4 • C and 60 • C afforded large aggregates with bimodal size distribution ( Figure S4). Thus, mCherry mRNA-encapsulated artificial viral capsid with unimodal size distribution can be obtained by incubation with the 1:9 co-assembly mixture at temperatures of 25 • C-37 • C. An aqueous solution of mCherry mRNA in PBS was added to the lyophilized powder of dT20-SS-β-annulus:β-annulus peptide followed by dilution with PBS and incubation for 30 min at 25 °C to prepare capsids containing mCherry mRNA. The electrophoretic mobility shift assay (EMSA) showed a band of mCherry mRNA significantly shifted by complexation with a co-assembly mixture of dT20-SS-β-annulus:β-annulus peptide at a 1:9 molar ratio ( Figure 4A, lane 4). In contrast, the other co-assembly mixtures showed only a slight retardation of mRNA ( Figure 4A, lane 5-7), which reflect almost neutral net charge and relatively small size of the β-annulus peptide. The migration position of a mixed solution of unmodified β-annulus peptide and mCherry mRNA was not shifted from mCherry mRNA alone ( Figure 4B). DLS and TEM images of the 1:9 co-assembly complexes with mCherry mRNA show the formation of spherical assemblies of size 52 ± 10 nm ( Figure 5A,B). We previously reported that the N-terminus of the β-annulus peptide is directed toward the interior of the capsid and the surface ζ-potential of the capsid is almost zero at neutral pH [25]. Thus, when negatively charged RNA is encapsulated at neutral pH, the artificial viral capsid should minimally migrate from the applied position due to the capsid charge shield. Therefore, EMSA results indicate that the dT20-SS-β-annulus:β-annulus peptide 1:9 co-assembly encapsulated mCherry mRNA via hybridization with dT20 directed to the interior of capsids. EMSA shows that mCherry mRNA could be complexed with the 1:9 co-assembly at 4 °C, 25 °C, 37 °C, and 60 °C ( Figure S3). However, complexation at 4 °C and 60 °C afforded large aggregates with bimodal size distribution ( Figure S4). Thus, mCherry mRNA-encapsulated artificial viral capsid with unimodal size distribution can be obtained by incubation with the 1:9 co-assembly mixture at temperatures of 25 °C-37 °C.

Nuclease Resistance of mRNA-Encapsulated Artificial Viral Capsid
Resistance of mCherry mRNA-encapsulated artificial viral capsid to endoribonuclease (RNase A) was confirmed by EMSA. Naked mCherry mRNA was digested by RNase A within 10 min (Figure 6(Aa),B lane 3). Conversely, mRNA in artificial viral capsids (band near the loading well) was minimally digested after 60 min ( Figure 6(Ac),B lane 5-8) due to protection by the capsid. For the artificial viral capsids modified with dT 20 , the exteriors were constructed as a control by self-assembly of β-annulus peptide modified with dT 20 at the C-terminal that directed dT 20 to the exterior of the capsid. The mixture of mCherry mRNA and the dT 20 -modified capsid at the exterior did not cause a significant mobility shift ( Figure 6B lane 9), indicating that mRNA was not encapsulated. mRNA was rapidly digested by RNase A within 10 min (Figure 6(Ac),B lane 10). Thus, mRNA is protected from degradation by artificial viral capsids.

Nuclease Resistance of mRNA-Encapsulated Artificial Viral Capsid
Resistance of mCherry mRNA-encapsulated artificial viral capsid to endoribonuclease (RNase A) was confirmed by EMSA. Naked mCherry mRNA was digested by RNase A within 10 min ( Figure  6(Aa),B lane 3). Conversely, mRNA in artificial viral capsids (band near the loading well) was minimally digested after 60 min ( Figure 6(Ac),B lane 5-8) due to protection by the capsid. For the artificial viral capsids modified with dT20, the exteriors were constructed as a control by self-assembly of β-annulus peptide modified with dT20 at the C-terminal that directed dT20 to the exterior of the capsid. The mixture of mCherry mRNA and the dT20-modified capsid at the exterior did not cause a significant mobility shift ( Figure 6B lane 9), indicating that mRNA was not encapsulated. mRNA was rapidly digested by RNase A within 10 min ( Figure 6(Ac),B lane 10). Thus, mRNA is protected from degradation by artificial viral capsids. Gel electrophoresis assay of digestions mRNA by RNase A at 37 °C for 10 min (mRNA alone and β-annulus-dT20) and 10-60 min (dT20-SS-β-annulus:β-annulus = 1:9 co-assembly). The gel was stained with GelRed.

In-Cell Expression of mCherry mRNA Encapsulated in Artificial Viral Capsid
Uptake of mCherry mRNA-encapsulated capsids into human hepatoma HepG2 cells were analyzed by CLSM to evaluate mCherry mRNA expression inside cells. mCherry mRNA and a strong transfection reagent (TransIT-mRNA) were incubated with the cells for 48 h at 37 °C as a positive control; significant red fluorescence of mCherry was observed (Figure 7(Aa)). Unfortunately, minimal expression of encapsulated mCherry mRNA was observed as similar to the naked mCherry mRNA( Figure 7A(b,c)). We employed a histidine 16-mer (His16), a cell-penetrating peptide uncharged under physiological conditions [36,37] to enhance cell permeability of capsids. We

In-Cell Expression of mCherry mRNA Encapsulated in Artificial Viral Capsid
Uptake of mCherry mRNA-encapsulated capsids into human hepatoma HepG2 cells were analyzed by CLSM to evaluate mCherry mRNA expression inside cells. mCherry mRNA and a strong transfection reagent (TransIT-mRNA) were incubated with the cells for 48 h at 37 • C as a positive control; significant red fluorescence of mCherry was observed (Figure 7(Aa)). Unfortunately, minimal expression of encapsulated mCherry mRNA was observed as similar to the naked mCherry mRNA( Figure 7A(b,c)). We employed a histidine 16-mer (His16), a cell-penetrating peptide uncharged under physiological conditions [36,37] to enhance cell permeability of capsids. We constructed artificial viral capsids modified with His16 on the exterior and using the 1:9 co-assembly of dT 20 -SS-β-annulus peptide and β-annulus peptide modified with His16 at the C-terminal (β-annulus-His16). Significant red fluorescence of mCherry was observed in cells incubated with His16-modified capsids (Figure 7(Ad)). The expression level of mCherry was quantified by fluorescence intensity in the cell areas of the CLSM image using the software ImageJ. The expression level using His16-modified capsids was 26.5 times the expression observed with unmodified capsids ( Figure 7B). This indicates that the intracellular expression of mCherry was significantly enhanced by the modification of His16 on the surface of the artificial viral capsid. We have demonstrated in our previous work that the reductive cleavage of the disulfide bond of the artificial viral capsid by dithiothreitol (DTT) caused the controlled-release of ssDNA with the destruction of the capsid [35]. mCherry mRNA was likely released from capsids in the reducing environment in cells.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 8 of 11 annulus-His16). Significant red fluorescence of mCherry was observed in cells incubated with His16modified capsids (Figure 7(Ad)). The expression level of mCherry was quantified by fluorescence intensity in the cell areas of the CLSM image using the software ImageJ. The expression level using His16-modified capsids was 26.5 times the expression observed with unmodified capsids ( Figure 7B). This indicates that the intracellular expression of mCherry was significantly enhanced by the modification of His16 on the surface of the artificial viral capsid. We have demonstrated in our previous work that the reductive cleavage of the disulfide bond of the artificial viral capsid by dithiothreitol (DTT) caused the controlled-release of ssDNA with the destruction of the capsid [35]. mCherry mRNA was likely released from capsids in the reducing environment in cells.

Conclusions
We constructed artificial viral capsids co-assembled from viral β-annulus peptides bearing dT 20 at the N-terminus and unmodified peptides. Capsids bearing interior dT 20 enable encapsulation of mCherry mRNA bearing poly(A) tails via hybridization to form ribonuclease resistant spherical assemblies of approximately 50 nm in diameter. Further, we demonstrate novel material for mRNA delivery by His16-modification on the exterior of capsids. Further research will demonstrate potential applications for artificial viral capsids as delivery systems for nucleic acid drugs.