Size Dependent Cellular Uptake of Rod-like Bionanoparticles with Different Aspect Ratios

Understanding the cellular internalization mechanism of nanoparticles is essential to study their biological fate. Especially, due to the anisotropic properties, rod-like nanoparticles have attracted growing interest for the enhanced internalization efficiency with respect to spherical nanoparticles. Here, to elucidate the effect of aspect ratio of rod-like nanoparticles on cellular uptake, tobacco mosaic virus (TMV), a typical rod-like bionanoparticle, is developed as a model. Nanorods with different aspect ratios can be obtained by ultrasound treatment and sucrose density gradient centrifugation. By incubating with epithelial and endothelial cells, we found that the rod-like bionanoparticles with various aspect ratios had different internalization pathways in different cell lines: microtubules transport in HeLa and clathrin-mediated uptake in HUVEC for TMV4 and TMV8; caveolae-mediated pathway and microtubules transport in HeLa and HUVEC for TMV17. Differently from most nanoparticles, for all the three TMV nano-rods with different aspect ratios, macropinocytosis takes no effect on the internalization in both cell types. This work provides a fundamental understanding of the influence of aspect ratio on cellular uptake decoupled from charge and material composition.

Based on these concerns, we report the study of the effect of aspect ratio on cellular uptake mechanism of rod-like bionanoparticles in epithelial and endothelial cells. As shown in Fig. 1, TMV can be broken into short rods by ultrasonic treatment. With this method, these short rods can still maintain their original rod morphology, exactly the same diameter, identical surface property and material compositions. Short rods with different aspect ratios can be separated by sucrose density gradient centrifugation separation. These bionanoparticles with different aspect ratios were then used to study their specific uptake mechanism in both epithelial and endothelial cells.

Results and Discussion
Ultrasound could generate acoustic cavitation in liquids, which contains the formation, growth, and implosive collapse of bubbles. These processes produce extremely high temperatures and pressures in a microscopic region, thus trigger reactivity 34 . In another hand, viruses are generated by the nature own design with the protein assembly by the weak interaction. Thus, the mechanical force generated by the ultrasonication could easily destroy the integrity of TMV. In a typical experiment, a 5 mg/mL TMV in 0.01 M pH 8.0 phosphate buffer was treated with ultrasound in the ice bath for 30 minutes with three times. After treatment, the sample was directly characterized by the transmission electron microscopy (TEM). As shown in Fig. 2b, comparing to the TMV with the length 300 nm (Fig. 2a), after ultrasonication, short rods with the length 20~200 nm (coded as usTMV) are visualized. The diameter of short rods still remains 18 nm, which is the same as native TMV.
One critical consideration is the structural stability of TMV after ultrasonic treatment. Sonication could denature proteins through large pressure, temperature gradients, high shear forces and generating free radicals 35,36 . However, the suitable ultrasound parameter, such as lower ultrasound power and shorter irradiation time, could minimally damage protein 37 . Based on that, we chose thrice 30 min ultrasonic treatment in ice bath. In our research, the main focus is the aspect ratio-dependent internalization mechanism of rod-like nanoparticles, which have the monodispersed diameter and behave the same physiochemical nature. Whether the TMV short rods still have bio-activity after ultrasonication treatment is very important. TMV coat proteins have the capacity to self-assemble into well-defined nano-rods or nano-disks at specific pH value and ionic strength 38,39 . So we investigate the bio-activity of TMV short rods through their assembling behavior. As shown in Fig. 2c,d, ultrasonic treatments have nearly no effect on the self-assembly performance of TMV short rods: short rods can still reassemble into long rods in pH 5.5 acetate buffer, subsequently disassemble into short rods in pH 8.0 phosphate buffer. Furthermore, the assembly and disassembly behavior of TMV short rods is a reversal process. Based on above analysis, ultrasonication did not alter the same bioactivity of short rods and long rods.
Apart from the structural stability, another important item is if the surface charge of TMV short rods is changed after ultrasonication, which is one of the factors affecting cellular internalization. We measured the zeta potentials of TMV in PBS before and after ultrasonication (Table 1), and the results showed that ultrasonication did not change the surface charge in pH 7.4 (physiological pH) and pH 8.0 (experimental pH) PBS. Since the shape, bioactivity and surface charge are steady, we can treat this work as merely size-dependent study.
To maintain the outer surface properties, we labeled TMV with Rhodamine B (RB) onto its inner cavity through one step reaction between amino groups of RB and two carboxylic groups of TMV to render it fluorescence detectable in the cellular uptake experiments (Fig. 3a). After reaction, the TEM showed that TMV still kept intact (Fig. 3b). The SDS-polyacrylamide gel electrophoresis (SDS-PAGE) demonstrated that TMV was successfully labeled by RB (Fig. 3c): the existence of red emission bands meant that RB was chemically conjugated onto TMV inner cavity. The number of RB labels per virion was calculated based on the UV-Vis spectroscopy. Through absorbance at 540 nm to confirm RB concentration, together with the Modified Lowry Protein Assay Kit to determine the exact TMV concentration, the conjugation density of RB was calculated as 0.4 RB molecules per TMV coat protein (852 dye molecules per TMV particle). Fluorescence spectra turned out that the conjugated RB provided TMV remarkable fluorescence emission (Fig. 3d).
After labeling, RB-TMV was treated with ultrasonication in the same condition as described before. The mixture of short rods was then separated by the sucrose density gradient centrifugation technique. This technique has been proved to be an efficient method to separate nanorods with different aspect ratios. As nanorods with different lengths have different sedimentation rates, giving rise to a broad band in the centrifuge tube under the centrifugal velocity, we collected the fraction from different part of the centrifuge tube: the upper part was the rods with short aspect ratios; the lower part was the rods with long aspect ratios. These rods were characterized by TEM and dynamic laser scattering (DLS) measurements. Comparing to RB-TMV with length 300 nm (Fig. 4e) and a hydrodynamic diameter of 80.3 nm, the short rods with average length about 70 nm (Fig. 4a) can be clearly visualized. DLS measurement shows the short rods have a narrow size distribution with hydrodynamic diameter of 19.4 nm (Fig. 4b). Collected from the lower part, the rods with average length about 140 nm (Fig. 4c) and narrow size distribution (measured by DLS, Fig. 4d) were obtained.  By co-culturing as-prepared RB labeled TMV rods with different cell lines for determined time, the cells were visualized on confocal laser scanning microscopy (CLSM). Wege and Steinmetz group recently reported that the behavior of TMV, RGD-labeled TMV and PEGylated TMV in epithelial tumor cells (HT-29 cells) and immune cells significantly rely on the aspect ratio 32 . However, the internalization mechanisms are not emphasized. For most nanoparticles, the transendothelial and transepithelial pathway are different, because of the difference in membrane proteins, permeability, cell polarity and membrane asymmetry between endothelial and epithelial cells. For example, the asymmetry of the cell membrane is more prominent in most epithelial cells; the caveolae expression on endothelial cell membrane is higher than that on epithelial cell membrane. So in our study, an epithelial cell line (HeLa cell line) and an endothelial cell line (human umbilical vein endothelial cell (HUVEC)) were used to investigate the endocytic processes of the TMV rods with different aspect ratios. As shown in Fig. 5, CLSM images showed that all the samples could be endocytosed into Hela cells and HUVEC cells, accumulating around cell nuclei. To evaluate the kinetics of particle internalization, we applied flow cytometry and studied the time course of cellular uptake from 15 min to 4 h (Fig. 6). Computer simulations have shown that cell membrane wrapping of anisotropic particles is intricate interplay between shape, initial orientation, surface adhesive energy, volume and aspect ratio [40][41][42][43][44][45][46] . Our data showed that the cellular internalization presented apparent dependence on aspect ratios of TMV. The internalization rate of TMV 4 and TMV 8 displayed cell type independent, and was obviously higher than TMV 17 . Shorter ones were found to undergo endocytosis easier than longer ones. The reason may be that the larger volume, increased surface area and high aspect ratio are unfavorable for complete membrane wrapping 40,41 . Especially, the uptake rate of TMV 4 was as high as 90% during the first 15 minutes. This high internalization efficiency of particles with aspect ratio of around 4 is in agreement with recent literature 9,13 . For the high-aspect-ratio TMV 17 , its internalization process was much slower, and its internalization kinetics exhibited different in various cell types: TMV 17 was ingested faster in HUVEC than in HeLa cells.
Furthermore, we choose the widely applied pharmacological inhibitors to confirm the occurrence of macropinocytosis, microtubules transport, clathrin-mediated endocytosis and caveolae-mediated endocytosis 47,48 . The inhibitory functions of the pharmacological inhibitors 17,49 and concentration used are listed in Fig. 7a. At the chosen concentration in Fig. 7a, the cells can achieve minimum 90% viability during 7 h co-incubation. The uptake efficiency of TMV 4 , TMV 8 and TMV 17 in HeLa cells and HUVEC cells were quantified by flow cytometry, shown in Fig. 7b-g. Amiloride blocks macropinocytosis, which is a non-specific pathway and is involved in most cell types. While for all the three TMV nano-rods with different aspect ratios, amiloride showed no inhibiting effect on both epithelial cell type (HeLa) and endothelial cell type (HUVEC). Clathrin-mediated endocytosis (blocked by chlorpromazine) is a best-characterized receptor-mediated pathway. It shows that it is a major internalization mechanism for both TMV 4 and TMV 8 in HUVEC cell line (about 50% decreasing, p < 0.01), and for all the three TMV rods, it plays a more important part in HUVEC cell line than in HeLa cell line. Caveolae, blocked by genistein, are a special type of lipid raft in many cell types, especially in endothelial cells. In Fig. 7, it shows that the caveolae-mediated mechanism behaves as a major internalization mechanism for all the three TMV rods in HUVEC cell line, and plays an important part for TMV 17 in HeLa cell. Nocodazole is a pharmacological inhibitor for microtubule polymerization, which is responsible for a variety of cell movements, including the intracellular transport and positioning of membrane vesicles and organelles. Figure 7 showed that it could reduce the uptake efficiency of TMV 17 to about 60% (p < 0.01) in both HeLa and HUVEC cell lines, and it was a significant inhibitor for TMV 4   Overall, the internalization mechanism of TMV nano-rods in both HeLa and HUVEC cells shows size-dependent. In HeLa cells (epithelial cell type), the prominent uptake mechanisms are caveolae-mediated endocytosis and microtubules transport for TMV 17 , microtubules transport for TMV 8 , and clathrin-mediated endocytosis and microtubules transport for TMV 4 . In HUVEC cells (endothelial cell type), the major internalization mechanisms are caveolae-mediated endocytosis and microtubules transport for TMV 17 , clathrin-mediated endocytosis and caveolae-mediated endocytosis for both TMV 8 and TMV 4 .

Conclusion
Bionanoparticle can be a great platform to act as a typical model to study the size or morphology dependent cellular internalization pathways. In this work we have shown that TMV based bionanorods with different aspect ratios have different pathways entering into different cells. Microtubules transport was mainly used in HeLa and clathrin-mediated uptake was mainly used in HUVEC for both TMV 4 and TMV 8 , while caveolae-mediated and microtubules transports were predominantly used in HeLa and HUVEC for TMV 17 . Specifically, we have shown that none of these protein nanorods enters into different cells by macropinocytosis, which is dramatically different with synthetic nanorods. Additionally, the kinetic studies have shown that TMV short rods with aspect ratio of 4 and 8 are ingested much faster by both cells than TMV with the aspect ratio of 17. Our studies may have profound significance for the design of bionanoparticles based drug or gene carrier system.

Experimental Section
Materials and Chemicals. All reagents were used as received from commercial sources. Rhodamine B isothiocyanate (RBITC) and all the pharmacological inhibitions were purchased from Sigma. Ethylenediamine (EDA) and N-(3-Dimethylaminopropyl)-N'-ethylcarb odiimide (EDC) were obtained from Aladdin. Labeling TMV with RBITC-EDA. TMV was extracted from infected leaves using established method 23,50 .
The concentration of TMV was determined by Lowry Protein Assay kit (Solarbio Science & Technology Co., Ltd). A solution of 2 mg/ml TMV in HEPES buffer (pH 8.0, 10 mM) was added to the round-bottle flask with gentle vortexing. Then add 25 eq HoBt and 5 eq EDC to the flask successively every 30 min to activate the carboxyl groups in the inner cavity of TMV. Subsequently, put 5 eq RBITC-EDA dropwise into the flask. The mixture was stirred in dark at 4 °C for 18 h, during which add 5 eq EDC twice to the reaction system. Samples need to   Separation of TMV with Different aspect ratios. Stepwise gradients were prepared in 13 ml tubes by layering 2 ml volumes of 33% (wt/v), 30%, 27%, 24%, and 21% sucrose solutions diluted in PBS (pH 8.0, 10 mM) from bottom to top layer. The sucrose solutions were stored at 4 °C before using. Immediately, 0.2 ml usTMV in PBS was layered on top of the gradient. The tubes were centrifuged at 30,000 rpm for 2.5 h at 4 °C using a Beckman SW40 rotor. Each fraction (1 ml) was collected from top to down. The third fraction was collected as the short aspect ratio of TMV fraction, and the sixth fraction was collected as the medium aspect ratio of TMV fraction. While for the separation of long aspect ratio of TMV, the sucrose gradients were 70% (wt/v), 60%, 50%, 40%, and 30% sucrose in PBS from bottom to top, and 0.2 ml TMV was put on the top layer, then the same ultracentrifugation protocol was used. The sucrose layer with sample was collected as the long aspect ratio of TMV. After centrifugation, the TMV was purified from sucrose by dialysis against pH 8.0 PBS (10 mM) buffer in dark at 4 °C over 6 cycles (6 + hours each time) and condensed using a Microcon (Millipore) spin filter unit (MWCO 10 kD & 100 kD).
The hydrodynamic diameter of TMV with different aspect ratios in PBS was determined by Dynamic Light Scattering (DLS) on a DynaPro NanoStar instrument (Wyatt Technology) with a He-Ne laser (λ = 659 nm) operated at 10 mW and analyzed with Dynamics software v6.11. Experiments were run at 25 °C in triplicate, with 10 acquisitions per measurement.
The visualization of TMV was performed with a transmission electron microscopy (TEM, JEM-2100F JEOL, Japan) with an accelerating voltage of 120 kV and the samples were stained by 0.2% uranium acetate before observation.
Cell Culture. HeLa (human cervical cancer epithelial cell) and HUVEC (human umbilical vein endothelial cell) were incubated in DMEM medium with 1% antibiotics (penicillin-streptomycin) and 10% FBS at 37 °C in a humidified atmosphere containing 5% CO 2 .

Co-incubation of TMV with various aspect ratios with cells. HeLa epithelial cells and HUVEC
endothelial cells were seeded respectively onto glass coverslips in a 6 well culture plate overnight. Thereafter, the TMV with various aspect ratios dispersed in the culture medium and the cells were incubated for another 4 h. After removing the medium and washing thrice with PBS, the cells were fixed for 10 min with 4% paraformaldehyde in PBS at room temperature, and washed thrice with PBS. Then the cells were permeablized with 0.1% Triton X-100 in PBS for 5 min at room temperature, and washed thrice with PBS. F-actin was then stained with Alexa Fluore-488 phalloidin in 1% BSA/PBS for 15 min at room temperature, washed thrice with PBS and the nucleus was stained with To-PRO-3 iodide in PBS (10 ug/ml RNase/PBS) for 20 min at room temperature. Coverslips were washed thrice with PBS and water, and sealed using Pro Long-Gold antifade reagent. Cells were observed by Nikon Eclipse Ti confocal laser scanning microscopy (CLSM).

Pharmacological Inhibition Studies by Flow Cytometry.
For the performance of flow cytometry, HeLa epithelial cells and HUVEC endothelial cells were seeded (50,000 cells per well) in a 24 well tissue culture plate and then allowed to adhere overnight. Pharmacological inhibitors were used and added to cells at the concentration we used before 30 described in Table 1. Cells were incubated for 1 h after which TMV with various aspect ratios were added to cells (10 ug/ml) for another 4 h. Then cells were washed thrice with PBS, trypsinized and collected into 10% FACS buffer (10% FBS in PBS), and analyzed using a flow cytometry (BD Calibur) and 10,000 gated events were recorded. The untreated cells were taken for background fluorescence, which was subtracted from test samples. For each sample (TMV 17 , TMV 8 or TMV 4 ), cells without inhibitor pretreating were taken as the control group (set as the 100% uptake efficiency). Data show average and standard deviation from three independent experiments.
Kinetics of TMV internalization. The HeLa and HUVEC cell lines were used to investigate the uptake rate of TMV with various aspect ratios. Samples were incubated with cells over a time course ranging from 15 min to 4 h (37 °C, 5% CO2). After cell/particle incubation, the cells were washed and detached by trypsinization. Cells were then collected in PBS with 10% FBS, and samples were analyzed by flow cytometry. There were 10,000 cells measured in each sample.