Boron-incorporating hemagglutinating virus of Japan envelope (HVJ-E) nanomaterial in boron neutron capture therapy

ABSTRACT Combining immunotherapeutic and radiotherapeutic technique has recently attracted much attention for advancing cancer treatment. If boron-incorporated hemagglutinating virus of Japan-envelope (HVJ-E) having high membrane fusion ability can be used as a boron delivery agent in boron neutron capture therapy (BNCT), a radical synergistic improvement of boron accumulation efficiency into tumor cells and antitumor immunity may be induced. In this study, we aimed to develop novel boron-containing biocompatible polymers modified onto HVJ-E surfaces. The copolymer consisting of 2-methacryloyloxyethyl phosphorylcholine (MPC) and methacrylamide benzoxaborole (MAAmBO), poly[MPC-co-MAAmBO], was successfully synthesized by using a simple free radical polymerization. The molecular structures and molecular weight of the poly[MPC-co-MAAmBO] copolymer were characterized by nuclear magnetic resonance and matrix-assisted laser desorption ionization time-of-flight mass spectrometry, respectively. The poly[MPC-co-MAAmBO] was coated onto the HVJ-E surface via the chemical bonding between the MAAmBO moiety and the sugar moiety of HVJ-E. DLS, AFM, UV-Vis, and fluorescence measurements clarified that the size of the poly[MPC-co-MAAmBO]-coated HVJ-E, HVJ-E/p[MPC-MAAmBO], to be about 130 ~ 150 nm in diameter, and that the polymer having 9.82 × 106 ~ 7 boron atoms was steadily coated on a single HVJ-E particle. Moreover, cellular uptake of poly[MPC-co-MAAmBO] could be demonstrated without cytotoxicity, and the hemolysis could be successfully suppressed by 20%. These results indicate that the HVJ-E/p[MPC-MAAmBO] may be used as boron nanocarriers in a combination of immunotherapy with BNCT.


Introduction
Cancer immunotherapy has been approved as an advanced therapeutic technique, and the effectiveness of several immune checkpoint inhibitors for various types of cancer such as melanoma, nonsmall cell lung cancer, etc. has been proven [1][2][3]. Recently, a new immunotherapeutic strategy using inactivated Hemagglutinating Virus of Japan Envelope (HVJ-E) as a gene delivery vector has attracted much attention, CONTACT Takehiko Tsukahara ptsuka@lane.iir.titech.ac.jp Laboratory for Advanced Nuclear Energy, Tokyo Institute of Technology, 2-12-1-N1-6, Ookayama, Meguro-ku, Tokyo 152-8550 Japan because HVJ-E has high membrane fusion ability induced by fusion (F) and hemagglutininneuraminidase (HN) proteins on the surface [3][4][5]. Previous studies have found that the membrane fusion ability of HVJ-E activates the antitumor immunity by inducing interleukin (IL)-6 and C-X-C motif chemokine (CXCL) 10 expression in dendritic cells, and more than half of the tumors in mice can thus be destroyed [6,7]. Induction of tumor-selective apoptosis is also produced by the membrane fusion, since the viral RNA genome fragments of HVJ-E activate the retinoic acid-inducible gene-I (RIG-I) signaling pathway and upregulate the tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and Noxa [8]. Moreover, HVJ-E incorporating antitumor immunotherapeutic agents have been demonstrated to be useful as a drugdelivery carrier for cancer treatment [9][10][11][12]. However, there is a serious problem in that membrane fusion of HN proteins with erythrocyte on the HVJ-E surface often occurs, resulting in hemagglutination and hemolysis [13]. In order to overcome these disadvantages, surface coating of HJV-E which can minimize the hemagglutination and the nonspecific adsorption has been investigated by some groups including ourselves, and HVJ-E surfaces have been successfully coated with polymeric materials by means of cationized-gelatin conjugation and layer-by-layer (LbL) assembly methods [14][15][16].
Radiation stimulates cytotoxic T lymphocyte (CTL) activity, and leads to not only systemic antitumor immune response but also growth suppression of nonirradiated metastatic tumors at distant sites from irradiated primary tumor sites. This phenomenon called the abscopal effect has been known to be facilitated by radiotherapy in combination with immunotherapy [17][18][19]. Recent studies have revealed that the combination of immunotherapy with radiotherapy enables the synergistic enhancement of cancer-treatment efficacy, and global clinical trials for some types of cancer are ongoing [20]. Among the radiation therapy, boron neutron capture therapy (BNCT) is a powerful cancer cell-targeted radiation treatment. The nuclear reaction of thermal neutrons and boron-10 isotopes ( 10 B, 19.9% natural abundance) emits alpha (α) particles and lithium atoms because of the 10 B(n,α) 7 Li reaction. Since the range distance of the particles, ca. 10 µm, is quite close to the size of single cell, the selective destruction of target tumor cells can be accomplished without any effect on normal cells [21]. Low molecular weight boron agents such as sodium borocaptate (BSH) and p-boronophenylalanine (BPA) have been clinically used as boron-containing pharmaceuticals [22,23], while they exhibit a short-term retention time in tumor cells, and often cause vascular endothelial injuries because of the severe increase of boron concentrations in blood [23][24][25][26][27][28][29]. Various boronincorporated macromolecular agents with high boron contents such as liposomes, polymeric nanomicelles, and antibodies have also been produced, and their physiological-pathophysiological evaluation has been performed [30][31][32][33][34][35][36]. However, the macromolecular approaches are not satisfactory in terms of the insufficient boron accumulation efficiency in tumor cells, low boron content per unit weight, and the complicated synthesis.
For overcoming these disadvantages, Fujii et al. has applied HVJ-E to boron delivery carriers for BNCT treatment by encapsulating BSH into the cationized gelatin conjugate HVJ-E (CG-HVJ -E-BSH). The results have shown that CG-HVJ -E-BSH has both high tumor/normal liver cells 10 B concentration ratio and high retention in the tumor cells, and can radically enhance the therapeutic efficacy without normal liver injury as compared with BSH itself [15]. Accordingly, utilization of boronincorporating HVJ-E vector is expected to open the door of pharmaceutical breakthrough in BNCT treatment, since the vector induces the synergistic effect of combining drug delivery and immune activity. In spite of its great potential, there has been no other research on the boron-incorporating HVJ-E. The development of novel boron-labelled HVJ-E vector is also quite challenging.
If the boron-labelled HVJ-E can be used as a potential radioimmunotherapeutic agent for BNCT, a radical synergistic improvement of boron accumulation efficiency into tumor cells and enhancement of the abscopal effect may be induced. Moreover, it is expected that modification of boroncontaining polymers on HVJ-E surfaces can not only achieve high loading of boron atoms and hemolysis suppression but also provide anti-cancer therapeutic functions by encapsulating anticancer agents inside the surface-modified HVJ-E. Therefore, in this study, we aimed to synthesize by means of simple polymerization and surface modification techniques a biocompatible boron-containing copolymer-coated HVJ-E which allows to be a functional nanocarrier for BNCT.
The poly[MPC-co-MAAmBO] was modified onto HVJ-E surfaces, and the amount of boron atoms and the stability of the polymer coated on HVJ-E were evaluated by using UV-Vis and fluorescence spectroscopic analyses. Moreover, the effects of the poly [MPC-co-MAAmBO] on hemolysis, cellular viability, and cellular uptake were also examined.

Preparation of poly[MPC] and poly[MPC-co-MAAmBO]
MAAmBO (216.8 g mol -1 ) was synthesized as follows. 1.7~1.8 M NaOH aqueous solution was cooled down to below 1.0°C, and 1.0 g of 5-amino-2-(hydroxymethyl)phenylboronic acid cyclic monoester (6.7 mmol) was added and dissolved. After that, 0.98 mL of methacryloyl chloride (13.5 mmol) was very gently dropped into the solution during vigorous stirring, and the reaction was proceeded for 2 h under light shielding conditions. After a concentrated HCl solution was added slowly until pH equaled 2~3, pale-brown-colored MAAmBO precipitates could be recovered. The precipitates were filtrated and dried in vacuum overnight at 30°C.
Poly [MPC] and poly[MPC-co-MAAmBO] were synthesized by using free radical polymerization method. MPC (295.3 g mol -1 ) and MAAmBO as monomers and ACVA as a radical initiator were simultaneously added in a polypropylene tube. The amount of the added monomers and initiator is listed in Table 1. After 10 mL of ethanol degassed in an Ar gas bubbling was added in the reaction tubes, the monomers and initiator were dissolved. The solutions were heated up to 69°C and reacted for 24 h under Ar-filled glove box. After the polymerization, the resulting solutions were reduced in volume using a rotary evaporator, and dialyzed using regenerated cellulose membrane filters (Spectra/Pro 6, MWCO: 1 kDa, Spectrum Laboratories, Inc.) in methanol and distilled water. The dialyzed solution was lyophilized, and fine powder of poly[MPC] (white) or poly[MPCco-MAAmBO] (pale yellow) was successfully recovered.

Hemolysis test
About 5 mL of whole chicken blood was dispersed in 40 mL D-PBS(-), and centrifuged at 1000 rpm for 20 min and at 4°C. After the supernatant was removed, only the precipitate was suspended again in 40 mL D-PBS(-) for washing. By repeating this procedure three to five times, the chicken erythrocyte was thoroughly purified. 8% chicken erythrocyte suspension was prepared by mixing 4 mL of the purified erythrocyte precipitate in 46 mL D-PBS(-). The suspension was stored in a refrigerator. Each HVJ-E, HVJ-E/p[MPC], or HVJ-E/p[MPC-MAAmBO] sample (10 μL) with polymer concentrations of 1, 10, 25, 50, 100, and 200 mg mL -1 was mixed with 8 vol% chicken erythrocyte suspension (500 μL), and the samples were incubated at 37°C for 3 h. After incubation, each suspension was centrifuged at 2000 rpm for 5 min and at 4°C and the supernatant was recovered, and the absorbance of the recovered supernatant was measured at 542 nm for the evaluation of hemolysis (N = 3).

Evaluation of cell viability
HepG2 cells were seeded on a 96 well microplate (1.0 × 10 4 cells/well) in 200 μL FluoroBrite DMEM supplemented with 10% FBS and 1% Anti-Anti at 37°C under 5% CO 2 . After overnight incubation, the medium was exchanged to DMEM/D-PBS(-) (= 9:1) containing 1 mg mL −1 poly[MPC-co-MAAmBO], and the cells were incubated in the same condition for 24 h. Alamar blue assay was adopted to evaluate cellular viability. 20 μL of alamarBlue® reagent (Invitrogen, Carlsbad, CA, USA) was added in each well, and the cells were incubated again in the same condition for 2 h. The fluorescence intensity at 585 nm exited by 540 nm (N = 5) of the HepG2 cells treated with poly[MPC-co-MAAmBO] was measured, and compared with that of control HepG2 cells without poly[MPC-co-MAAmBO].   Since the 1 H-NMR spectrum was found to be quite similar to previous studies [40,41], the signals at around 2.0 ppm, 5.0 ppm, 5.5~6.0 ppm, 7.3~8.0 ppm, 9.2 ppm, and 10.0 ppm were assigned to CH 3 group in methyl methacrylate (MAAm) (peak; a), CH 2 group in BO (peak; g), CH 2 group in MAAm (peak; b), phenyl groups in BO (peaks; d, e, f), amide group (peak; c), and OH group (peak; h), respectively.

Results and discussion
We measured 11 B-NMR spectra of the synthesized MAAmBO in H 2 O/D 2 O (= 9:1) solvent at different pH (4.6 and 11.6) and at an ambient temperature. As shown in Figure 2(b), a broad peak at 27 ppm which was assigned to the boron atom of MAAmBO was observed for the case of pH = 11.6, while the peak is shifted to a lower magnetic field as a result of decrease in pH (pH = 4.6) and appears at 32 ppm. This result suggests that the boronic acid groups in MAAmBO exist as ionized >B(OH) 2 in a solution with higher pH, but for lower pH, the boronic acid is neutralized as >B-OH. The broad background and sharp peak at around 20 ppm are attributable to the borosilicate glass used as a sample tube and the boric acid produced by hydrolysis of BF 3 , respectively.  Table 2 shows the particle sizes of HVJ-E, HVJ-E/ p[MPC], and HVJ-E/p[MPC-MAAmBO] measured by DLS. The averaged diameter of all of the polymercoated HVJ-E was estimated as 110-150 nm. It seems that the particle size of the polymer-coated HVJ-E is smaller than that of natural HVJ-E (160 nm). Such size reduction of HVJ-E itself by polymer modification demonstrates an opposite tendency as compared to the previously reported polymer CG-HVJ-E-BSH [15]. These results indicate that the spherically structures of HVJ-E are maintained even after the polymer coating, while the effective particle sizes were decreased due to the hyper-hydrophilic biocompatible straight structure of HVJ-E/p[MPC] and HVJ-E/p [MPC-MAAmBO] in aqueous solutions.

Characterization of polymer-modified HVJ-E
The shapes of HVJ-E, HVJ-E/p[MPC], and HVJ-E/p [MPC-MAAmBO] nanoparticles were observed by AFM. Since the particle surfaces are hydrophilicity, the solutions containing the particles were dropped on hydrophobic glass surfaces modified by octadecyltrichlorosilane (ODS) vapor and slowly dried for suppressing corruption of the particles. Figure 5 Figure 5(d). We found that the averaged sizes of HVJ-E and HVJ-E/p[MPC] were about 150 nm which was quite consistent with DLS results. On the other hand, as seen in Figure 5(c-d), in the case of HVJ-E/p[MPC-MAAmBO], the aggregation and the structural deformation were generated under dry conditions because of the hyperhydrophilic structure, resulting in the averaged size of about 180 nm wide and 20 nm height. These results indicate that all of the HVJ-E species well maintained the spherical shapes with 100 nm sizes regardless of polymer modifications. Figure 6 shows UV-Vis absorption spectra of the supernatant solutions of HVJ-E/Cy5-p[MPC] and HVJ-E/Cy5-p[MPC-MAAmBO] at pH 7.6. The absorption maxima at the wavelengths of 598 and 642 nm, which were assigned to Cy5, were observed for both Cy5-modified HVJ-E polymers. Accordingly, we concluded that HVJ-E may be adequately modified by poly[MPC] and poly[MPC-co-MAAmBO] polymers at the concentration range of 100 mg mL -1 . Figure 7(a) shows fluorescence spectrum of HVJ-E/ Cy5-p[MPC-MAAmBO]. We observed not only a peak corresponding to excitation wavelength (642 nm) but also a peak assigned to Cy5 (657 nm, see the inset in Figure 7(a)), and the fluorescence intensity obtained at 657 nm was determined as 11 ± 3 (N = 4). Calibration curve of emission intensity vs. Cy5-poly[MPC-co-MAAmBO] concentration is shown in Figure 7 (b). The concentration of the HVJ-E/Cy5-p[MPC-MAAmBO] could be calculated as 2.44 × 10 -5 g mL −1 by using the calibration curve. From the calculated concentration, the amount of the poly[MPC-co-MAAmBO] per single HVJ-E particle could be determined as 1.36 × 10 -13~-14 g particles -1 . As already mentioned in the MALDI-TOF and the 1 H-NMR measurements, the poly[MPC-co-MAAmBO] (MPC = 295.3 g mol -1 , MAAmBO = 216.8 g mol -1 ) has the number-average molecular weight (M n ) of about 2500, and contains 32.4 mol% MAAmBO moiety. Therefore, the number of boron atoms could be estimated as 3.0 per a single poly[MPC-co-MAAmBO] copolymer molecule, resulting in 7.22 × 10 20 boron atoms in one gram of the copolymer. By multiplying the value by 1.36 × 10 -13~-14 g particles -1 of HJV-E, boron atoms of 9.82 × 10 6~7 were found to be incorporated on a single HVJ-E particle. This value approaches to previous BSHencapsulating HVJ-E with 3.72 × 10 10 boron atoms per particle which was prepared by loading BSH (6667 μg boron) into HVJ-E suspension (1.0 × 10 10 particles) [15]. Accordingly, the HVJ-E/Cy5-p[MPC-MAAmBO] nanoparticle which enables both high loading of boron atoms and hemolysis suppression is expected to be a promising nanocarrier for BNCT.
UV-Vis absorbance spectrum of D-PBS(-) solution containing the glucose-treated HVJ-E/Cy5-p[MPC-MAAmBO] was measured (N = 5), and compared with non-glucose-treated HVJ-E/Cy5-p[MPC-MAAmBO]. As listed in Table 3, the UV-Vis absorbance at 642 nm of glucose-treated and non-glucose-  showed almost the same. From the calculation of the t test, the t-value and the p value were determined as 0.164 and 0.874, respectively. This results verified that there was no statistically significant difference of the UV-Vis absorption between glucose-treated and nonglucosetreated HVJ-E/Cy5-p[MPC-MAAmBO] in D-PBS(-) solutions, and that effect of glucose on the stability of polymer coated on HVJ-E is negligible.  Table 4, we found that the cellular viability of poly[MPC-co-MAAmBO] case was determined as 107% which is consistent with that of control case. When t test was carried out for the difference of the cellular viability between with and without poly[MPC-co-MAAmBO], a p value of about 0.54 was obtained. This fact indicates that administration of 1 mg mL -1 poly[MPC-co-MAAmBO] generates no apparent cytotoxicity.

Evaluation of cellar uptake of poly[MPC-co-MAAmBO]
The fluorescence microscopic images of cellular uptake of Cy5-poly[MPC-co-MAAmBO] in HepG2 cells for 45 and 90 min are shown in Figure 8(a,b), respectively. The fluorescence color changes showed that significant cytoplasmic localization of the Cy5-poly[MPC-co-MAAmBO] was occurred within 45 min, and intracytoplasmic localization of the polymer inside cells was uniformly proceeded during incubation of 90 min. It is expected that such cellular uptake property of poly [MPC-co-MAAmBO] play an effective role for enhancing treatment effect of BNCT, since previous PHITS (Particle and Heavy Ion Transport code System) simulation has suggested that the intracellular  homogeneous distributions of 10 B agents is more effective than heterogeneous intra-and intercellular distributions for BNCT treatment [43].

Hemolytic activities of polymer coating
The hemolysis ratio was calculated from the changes in absorbance according to the following equation, where 'HVJ-E' and 'blank' are obtained from the absorbance of only HJV-E and buffer solution (pH 7.6), respectively.
Hemolysisratio % ½ ¼ Absorbance of sample À Absorbance of 00 blank 00 Absorbance of 00 HVJ À E 00 À Absorbance of 00 blank 00 Â 100 Figure 9(a,b) show the absorbance at 542 nm and the calculated hemolysis ratio of each supernatant solution    DLS and AFM measurements, the particle sizes of poly [MPC-co-MAAmBO]-coated HVJ-E (HVJ-E/p[MPC-MAAmBO]) with different concentrations were determined to be about 130~150 nm, slightly smaller than natural HVJ-E itself (160 nm). It was found that the spherical structure of HVJ-E is not degraded by polymer modification, while the hyperhydrophilic biocompatible straight structure of the poly[MPC-co-MAAmBO] generates a decrease in the effective particle sizes in aqueous solutions. Moreover, glucose tolerance of the poly[MPC-co-MAAmBO] on HVJ-E in a simulated blood solution was examined based on UV-Vis absorbance changes, resulting in that the polymer was steadily coated on the surface of the HVJ-E even in the environment of high glucose level.
In order to clarify biological and cytotoxic effects of the poly[MPC-co-MAAmBO], in vitro cellular viability and cellular uptake experiments were performed. The differences of fluorescence intensities of HepG2 cells treated with and without poly[MPC-co-MAAmBO] obtained by Alamar blue assay exhibited that the polymer has quite low cytotoxicity. The cellular uptake of Cy5-poly[MPC-co-MAAmBO] in HepG2 was examined by means of fluorescence microscopy. From the fluorescence color changes, intracytoplasmic localization of the Cy5-poly[MPC-co-MAAmBO] into the cells was found to proceed for 45~90 min.

Disclosure statement
No potential conflict of interest was reported by the authors.