Developing A Programmable, Self-Assembling Squash Leaf Curl China Virus (SLCCNV) Capsid Proteins Into “Nano-Cargo”-Like Architecture: A Next-Generation “Nanotool” For Biomedical Applications

A new era has begun in which pathogens have become useful scaffolds for nanotechnology applications. In this research/study, an attempt has been made to generate an empty cargo-like architecture from a high-profile plant pathogen of Squash leaf curl China virus (SLCCNV). In this approach, SLCCNV coat protein monomers are obtained efficiently by using a yeast Pichia pastoris expression system. Further, dialysis of purified SLCCNV-CP monomers against various pH strengthenened (5–10) disassembly and assembly buffers produced a self-assembled “Nanocargo”-like architecture, which also exhibited an ability to encapsulate the magnetic nanoparticles at in vitro. Bioinformatics tools were also utilized to predict the possible self-assembly kinetics and bioconjugation sites as well. The biocompatibility of “SLCNNV-CP-Nanocargo” particles was also evaluated by in vitro cancer cells, which eventually proved the particles to be versatile material for the next generation “nanotool” capable of housing various therapeutic or imaging agents.


Introduction
There has always been a high demand for biological entities which have the ability to act as a 61 multifaceted scaffold or template. Biological entities having intrinsic self-assembling properties 62 are recognized as an excellent aspect in nanotechnology applications [1][2][3][4]. The coat protein of 63 viruses are the kind of self-assembling peptides which form a fine architecture at a nanoscale 64 precision is called capsid. Recently viruses are being studied for their ability to self-assemble 8 150 Technologies, USA. Ferrous chloride hydrated extra pure (code no. 03846) and ferric chloride 151 anhydrous 98% extra pure (code no. 03817) were obtained from LobaChemie Pvt. Ltd,India. 152 Isolation and purification of virus particles 153 To investigate the assembling and disassembling processes, native Squash leaf curl China virus 154 (SLCCNV) was isolated from the natural host Benincasa hispida (winter melon), which was 155 collected from a field at Perambalur in the southern part of the state of Tamil Nadu in India. 156 Further, screenings and confirmations were made by molecular studies including PCR 157 amplification with coat protein-specific primers, which was already reported in our previous 158 work [37]. The virus particles or so-called viral nanoparticles (VNPs) were purified by 159 cesiumsulfate (Cs 2 SO 4 ) density gradient ultracentrifugation (CP100WX, HITACHI, JAPAN) 160 using fixed angle rotor P100AT2 (803,000xg) and swing bucket rotor P55ST2 (366,000 x g) with 161 necessary modifications [38]. Purified fractions of virus proteins were collected and dialyzed 162 against 0.1M phosphate buffer (pH 7.2) to remove the remnants of Cs 2 SO 4 . The purity of the 163 native virus particles was also confirmed through SDS-PAGE followed by a western blot 164 analysis [39,40]. For experimental control, the purified native virus particles were used all along 165 with SLCCNV-CP subunits to be expressed in yeast Pichia pastoris. 166 Heterologous expression of SLCCNV-CP in yeast Pichia pastoris 167 The coat protein expression comprises three principal steps: (a) insertion of the gene into an yeast Pichia pastoris GS115 strain was applied as a host cell to express SLCCNV coat protein 171 [41]. The plasmid pPICZαA containing methanol-inducible AOX1 promoter was used to 9 172 construct the full length SLCCNV coat protein gene [42]. The major advantage of using 173 pPICZαA (3.6 kb) vector for expression is the presence of a Zeocin-resistant gene for selection, 174 which has an alpha-factor secretion signal for directing the secreted expression of the 175 recombinant protein into the growth medium [43]. 177 To design an insert for expression, a complete coat protein gene (771bp) from the AV1 region of  of 1 ml/min. The eluted fractions containing the SLCCNV-CP were pooled and stored at -80˚C.

242
The obtained fractions were analyzed at 12% SDS-PAGE. Following electrophoresis, SDS-243 PAGE bands containing the protein of interest were excised from stained gels and subjected to 244 the in-gel trypsin digestion procedure as described elsewhere [47,48]. Spectral measurements  Investigation of self-assembly of SLCCNV ( native and Pichia-expressed) 259 In vitro assembly were performed by the conventional dialysis method (15kDa cut-off 260 membrane) against various pH assembly buffers. The assembly buffers with different pH (5 -261 10) were prepared by the addition of 1mM CaCl 2 [29,30,49,50]. In the first study, small volume 262 of native SLCCNV particles presence in the buffer of pH 7.2 (0.1M phosphate buffer) were 263 dialysis against the buffer of pH 5.0 (0.1M sodium acetate) at the cold temperature for 8 hours.

264
After that, the existing buffers in dialysis tank were exchanged to the other assembly pH buffers 265 of pH of 6.0 -8.0 (0.1M sodium phosphate), pH 9.0 (0.1M sodium borate) and pH 10.0 (0.1M 266 sodium glycine) at the stipulated time period of 8 hours in cold conditions. Eventually, the same 267 dialysis procedure were carried out for Pichia-expressed SLCCNV-CP subunits. Meanwhile DOE (http://nihserver.mbi.ucla.edu/Verify-3D/) structure evaluation server, which gives a visual 288 analysis of the quality of a putative crystal structure for proteins. The structural models were 289 validated using PROCHECK [45]. The query-modelled protein structure was submitted as PDB 290 files on the SAVES server (http://nihserver.mbi.ucla.edu/SAVES/) [53,54]. The quality of the 291 protein structure was validated from the Ramachandran plot through the Procheck server [55].

Prediction of assembly of SLCCNV-CP monomers 293
The PatchDock server is a simple and intuitive web interface, available at 294 http://bioinfo3d.cs.tau.ac.il/PatchDock [56]. The I-TASSER-derived monomeric structure of 295 SLCCNV coat protein was uploaded to the server and a docking request was submitted. Once the 296 prediction process was completed, the results were received by email containing the web link for 297 the predicted page. Modelling was performed to assemble a pentameric coat protein structure and 298 the it was repeated at three times. At the end of the process, pentameric structure of the 299 SLCCNV capsid protein was received. In addition, we tried to predict the solvent accessible 300 surface areas (SASA) by Accelrys Discovery Studio 2.5, because it was believed that one of the 301 key factors involved in the coat protein subunit assembly. Therefore, each modelled protein 302 structure (monomer to pentameric) was uploaded in Accelrys Discovery Studio 2.5 [57]. The

303
Connelly-type solvent accessibility helped in the identification of buried and exposed residues in 304 the solvent system. The solvent-exposed and buried residues were displayed as red and blue 305 colour, respectively. General biocompatibility or toxicity tests, aimed mainly at detection of the biological activity of 320 test substances, can be carried out in many cell types (e.g. fibroblasts, HeLa, hepatoma cells)

321
[59]. In this research/study, the biocompatibility feature of the expressed SLCCNV-CP alone 322 was determined in a dose-dependent manner by the MTT assay with A549 lung cancer cell line.

323
[60]. The A549 cells (1×104) were seeded onto 96 well flat-bottom culture plates and incubated 324 for 24 hours at 37˚C with 5% of CO 2 to reach the confluence in 70%. Over thirty various 325 concentrations of SLCCNV-CP (0.54 to18 μg/ml) were tested against the cell culture seeded in 326 the 96 well plate. The treated cells were kept incubated for 24 hours, after which 10µL of MTT 327 (5mg/ml in PBS) added to each well and the plate was incubated for another four hours at 37ºC.

328
The resulted formazan crystal was dissolved in 100µL of DMSO (Dimethyl sulfoxide) with a 329 gentle shake and the absorbance was measured at 595nm using an ELISA reader (BioTek power 330 wave XS, USA). All these steps were replicated at three times independently and the average has   This phenomenon was also confirmed after performing the confirmative assay of western blot 396 using an ACMV polyclonal antibody and an HRP-conjugated secondary antibody with DAB (3, 397 3'-diaminobenzidine) substrate. In the Figure 1H, the blotting membrane showed the bands 398 corresponding to the protein samples similar in the previous SDS-PAGE gel picture ( Figure 1G).

399
A single intensified band that is exactly the same size observed in the SDS-PAGE for native

441
Early investigation of disassembly and assembly of native SLCCNV particles through electron 442 microscopy, we found that there was no evidence of either particle disassembly or assembly   possibly being in an unfolding condition due to acidic buffer medium, which was also 465 determined by DLS hydrodynamic size measurement (Figure 4APanel 1-3). It was determined 466 that there particles of an average size of ~30 nm were present in the pH 5 buffer medium. We 467 also obtained a negative measurement value, which meant that size undetermined particles were 468 present in the buffer medium. When we observed an event of self assembly of the SLCCNV-CP 22 469 at buffer pH 6, the results in the HRTEM micrograph were surprising. In figure 4B Panel 2, the 470 electron micrograph shows a spectacular array made of coat protein monomers that can be a 471 disassembling event as was clearly visualised in the electron micrograph in the 100nm bar scale.

472
It has an average size of ~300nm which was also confirmed by DLS( Figure 4B Panel 1). We 473 confirm a speculation about the particular disassembling event of CP monomers at the pH of 6 474 after analyzing the results of buffer pH 7 further. However, at pH 7, a nanoscale structure that 475 collectively organized and assembled by CP monomers into a cargo-like architecture was 476 observed ( Figure 4C Panel 2). In the same micrograph, a single twin icosahedral structure similar 477 to the structure of Begomovirus (geminate) was also observed. The DLS data also confirmed that  Also, we observed aggregated population of CP monomers in the same electron micrograph. The 488 DLS for the same (pH 9) buffer medium also revealed that there were two populations of 489 aggregated particles with the size of ~100nm and ~700nm ( Figure 5B Panel 1). The CP 490 monomers in the buffer pH at 10, small-to-large population of protein aggregates were observed 491 in the size ranges at ~50nm and ~1100nm ( figure 5C Panel 2). Also, the DLS data were 23 492 confirmed the same, whereas 90% of protein particles were in the aggregated state at the buffer 493 pH 10 ( Figure 5C Panel 1). Overall, it was determined that the rate of self-assembly of 494 SLCCNV-CP was greater at the pH of 8. Finally, through DLS calculated mean hydrodynamic 495 diameters (D H ) for SLCCNV-CP in the presence of different pH were plotted against the 496 corresponding dialyzed pH buffers (Fig.S4.B, supplementary data).

497
The data for the self-assembly of SLCCNV-CP at varying pH with calcium ions point out that The protein structure was modelled using I-TASSER, a hierarchical approach to predict the 509 structure and function of an unknown protein from its sequence. The server generated the ten 510 best models of protein structure, from which the best one was selected based on the C-score 511 [64].The C-score of the predicted protein model was -2.45. The C-score of another predicted 512 model was significantly higher and hence model 1 was chosen for further study ( Figure 6A). The 513 overall quality of the predicted protein models was evaluated using the RMSD and the TM-24 514 score, which are 0.8±0.2 and 14.2±2, respectively. The I-TASSER method was used for the 515 prediction of protein structures with very little or no similarity, since the monomeric protein 516 structure was not available in any of the protein data repositories [45]. The predicted protein 517 model was validated on the SAVES PROCHECK server. The server constructed a 518 Ramachandran plot expressing the quality of the predicted protein model ( Figure 6B). The result 519 shows that 67% of the residues are found in the favoured region, while 24% of residues are in 520 allowed region and 9.1% of residues are in the outlier region. The 21 residues of Pro13, Arg19, 521 Phe23, Tyr27, Thr45, Arg51, Cys68, Arg80, Asp82, Ser113, Lys119, Asp123, Glu124, Asn125, 522 Lys127, His131, Phe159, Ser193, Phe202, Tyr219 and Asn238 were found in the outlier region.

523
The reduced quality of the predicted protein structure was due to the amino acid residues in the 524 outlier region. The less availability of similar protein templates resulted in the low quality of plot 525 [55].

527
The SLCCNV-CP subunits were docked to assemble the monomeric units into a pentameric 528 form of a viral capsid symmetry using a PatchDock online server. The docking method was 529 performed with an RMSD of 4.0. The docking of monomer-monomer, dimer-monomer, trimer-530 monomer and tetramer-monomer was carried out in four cycles to get a pentameric structure 531 ( Figure 6A).The structure was obtained with five subunits of monomer possessing secondary 532 structures for each subunit. The atomic contact energy (ACE) score was observed for each 533 subunit assembly (Table 1). The docking area increased with an increase in the number of 534 subunits, which has relatively decreased the contact energy resulting in the efficient binding of 535 two protein structures. The atomic contact energy score is based on the number of water contacts 536 replaced by the atomic contacts from the proteins. The atomic contact energy provides an 25 537 estimation of the free energy of the protein interactions. The lower ACE configuration suggests 538 lower free energy, which is more favourable for the formation of pentameric structure [65]. 539 Therefore, the pentameric assembly can be achieved by monomer and dimer intermediates [66].

540
Using Accelrys Discovery Studio 2.5. with Connelly-type modelling, the relative solvent 541 accessible surface areas (SASA) in the pentameric structure of coat protein could be clearly 542 figured out. In figure 6C, it shows that the predicted solvent accessible surface, which is 543 described in two colours (red and blue), helps in the identification of buried and exposed 544 residues. The results shows that there are hydrophilic residues covers the entire outer region of 545 capsomers and the hydrophobic regions are found in the deep core region. In this approach, SLCCNV-CP monomers were mixed with MNPs at disassembly buffer (pH 6) 577 and then dialyzed against assembly buffer (pH7). This showed prominent results after analysis 578 through UV-Vis spectroscopy, gel electrophoresis and HRTEM. Figure 9A, the UV-Vis 579 extinction spectra strongly represent the SLCCNV-CP monomers assembled and encapsulated 580 around the magnetic nanoparticles core. However, the original extinction peak for unmodified 581 MNPs changed to a low-intensity peak when encapsulated with SLCCNV-CP monomers, which 582 was reflected by the strong difference in the extinction peak. This reversible change in the 27 583 extinction spectrum clearly demonstrates that the MNPs are stabilized and might be dispersed by 584 the particle in solution. That condition is entirely due to decreased photon absorption by 585 SLCCNV-CP-nanocargo with MNPs. Encapsulation also confirmed at protein gel 586 electrophoresis without staining. In Figure 9B lane 1, the visible band shows that the control 587 unmodified MNPs were not separated. Figure

604
A key challenge in the biomedical field is the development of a smart delivery vehicle with an 605 ability to selectively entrap therapeutic and imaging molecules and display multiple