Virus‐Surface‐Mimicking Surface Clustering of AuNPs onto DNA‐Entrapped Polymeric Nanoparticle for Enhanced Cellular Internalization and Nanocluster‐Induced NIR Photothermal Therapy

Virus‐surface‐mimicking decoration of deoxyribonucleic acid (DNA)‐entrapped polymeric nanoparticle with AuNPs is demonstrated to lead to enhanced cellular uptake, improved gene transfection, and particularly efficient near‐infrared photothermal therapy that cannot be achieved by both of them separately. This hybrid nanosystem represents a novel paradigm of multipurpose organic–inorganic nanoplatform, especially for cancer treatments.


DOI: 10.1002/advs.201500108
decorating gene-entrapped polymeric nanoparticles with clustered gold nanoparticles (AuNPs) (Figure 1 ). This bio-mimic design has demonstrated signifi cantly enhanced cellular internalization of DNA payloads and consequently up to 100-fold promotion of transfection effi cacy. Inorganic nanoparticles, particularly AuNPs, have become a hot research area in the last decade; however, most of the reported studies are focused on exploiting them as the nanomatrix to undergo surface modifi cation with hydrophilic polymers or biomolecules for biomedical applications such as in vivo imaging and photothermal therapy. [ 13,14 ] Differently, the strategy of using AuNPs to modify organic nanosystem has nearly not been described. More attractively, this hybridization strategy paves a special avenue to realize the from-no-to-yes hyperthermia induction of AuNPs in the near-infrared (NIR) region. As known, one significant challenge for AuNP-induced photothermal therapy is that AuNPs, particularly spherical ones, mainly absorb light in the visible range with a more shallow penetration depth in tissue as compared to the therapeutic window in the NIR region. Several studies attract our interest that the AuNP aggregates self-assembled on cell surface can effectively increase the photothermal effi cacy under NIR irradiation. [15][16][17] It is supposed that the 3D location of AuNP clusters on the surface of DNA/ vector nanocomplex can produce this exciting feature, as actually proved herein with high effi ciency to kill cancerous cells, in contrast to the failure of both of them separately. This hybrid nanosystem represents a novel paradigm of gene-based multipurpose nanoplatform and can be extended to the convenient engineering of many hybrid inorganic-organic nanoplatforms with versatile multifunctions.
In a proof-of-principle experiment, a positively charged nanocomplex of plasmid pGL-3 condensed with lowly toxic polyethylenimine (PEI1800, M w = 1800 Da) was prepared at the optimal transfection N:P ratio of 20:1 and was used as the organic nanomatrix for the sequent fabrication of hybrid NPs. The nanocomplex possessed the surface zeta potential of +27 mV and the mean hydrodynamic diameter of 110 nm, as determined by dynamic light scattering (DLS; Figure S1, Supporting Information). Citrate-coated AuNP (particle size »10 nm, zeta potential approximately -35.8 mV) was slowly added into the nanocomplex solution under vibration. Driven by electrostatic interaction, the nanocomplex would readily be covered by a high density of smaller AuNPs (Figure 1 ). To simplify the research, the mass ratio of AuNP versus nanocomplex was optimized and This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
Gene therapy has shown great potency for treating various diseases, such as cancers. [ 1 ] One urgent task for gene therapy is focused on how to enable effi cient transport of therapeutic genes into host cells, for which the assistance of suitable vectors is required. [2][3][4] Nowadays, polycationic vectors have been among the most promising nonviral candidates for gene delivery, [ 5 ] but their transfection activities seem far from satisfactory and need to be improved. Because the cell entry of DNA/vector nanocomplex is one of the crucial factors affecting the overall transfection performance, universal strategies that can promote the cellular uptake are highly desired. [ 6 ] Up-to-date advance in understanding the structure-function relationship of enveloped viruses offers useful information to guide the design of artifi cial vectors, which opens possibilities of improving the transmembrane ability to a high level. [ 7 ] Many viruses such as infl uenza virus, herpes simplex virus (HSV), [ 8 ] and human immunodeficiency virus (HIV) [ 9 ] are identifi ed to possess a rough surface patched by glycoprotein spikes, as shown in Figure 1 . The nanoscale roughness of viral surface caused by glycoprotein spikes is thought to be "friendly" to cellular membrane and benefi t the cell entry. [ 10,11 ] It is further found that the sparsely distributed glycoproteins spikes would voluntarily cluster together to facilitate the viral entry into cells. [ 12 ] Drawing lessons from the viral topography, the present work reported a virus-surface-mimicking nanotechnology by fi xed at 7:1 throughout the study, based on the in vitro transfection experiment in HeLa cells (as discussed later). Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) strongly evidenced that AuNPs can bind to the nanocomplex and cluster into a rough periphery. TEM image showed that parent PEI1800/DNA nanocomplex displayed a regularly spherical shape with smooth surface and the particle size was around 50 ± 3 nm in dry state ( Figure 1 , inset (c)). In comparison, the addition of AuNPs led to a marked change of the size and topography of the nanoparticles. As seen from the typical SEM image, the formed PEI1800/pGL-3/AuNP nanoparticles still remained individually dispersed whereas possessing an apparently rough surface ( Figure 1 , inset (b)). The magnifi ed TEM image obtained at a lower dosage of AuNPs offered information to better understand the morphological changes, showing that the roughened surface was ascribed to the attachment of lots of AuNPs on nanocomplex ( Figure 1 , inset (a), small black dots on nanoparticles surface). As a result, the particle size was increased to about 200 ± 12 nm, which was slightly smaller than the hydrodynamic diameter of 225 ± 11 nm. Meanwhile, AuNP attachment induced a sharp decline of zeta potential from 27.0 ± 1.0 mV to 5.0 ± 0.4 mV (Table S1, Supporting Information), reconfi rming the successful manufacture of hybrid PEI1800/pGL-3/AuNP nanoparticles. Upon AuNP addition, the enhanced hydrodynamic dia meter of PEI1800/DNA naocomplex might associate with the charge neutralization occurring between the nanocomplex and the opposite charged AuNPs. As a consequence, the nanocomplex may be less compact than that without AuNP decoration.
To verify the feasibility of our virus-surface-mimicking idea, the intracellular internalization and the in vitro transfection of the AuNP-absent and -present nanocomplexes were comparatively explored. Four types of cell lines including human cervical cancer (HeLa) cells, monkey kidney fi broblast (COS7) cells, human hepatoma (HepG2) cells, and mouse fi broblast (3T3) cells were tested to evaluate the universal adaptability of this strategy. Confocal laser scanning microcopy (CLSM) provided an intuitive inspection over the cellular uptake of YOYO-1-labeled plasmid pGL-3. As shown in Figure 2 A, the green fl uorescence with strong intensity appeared in all the cell lines after the cells were exposed to PEI1800/pGL-3/AuNP for 4 h. In comparison, the treatments with AuNP-absent nanocomplex gave much lower cell-entry effi ciencies of entrapped DNA, as refl ected by the considerably weakened fl uorescence inside cells. Such an entry promotion caused by AuNP decoration appeared to be general, regardless of the investigated cell types. Moreover, quantitative analysis using fl uorescence-activated cell sorting offered solid data regarding the ability of delivering pGL-3 into cells. In consistence with the CLSM observation, the obtained profi les showed that the amount of YOYO-1-positive cells when exposed to PEI1800/pGL-3-YOYO-1/AuNP was at a much higher level than that treated with the counterpart of PEI1800/pGL-3-YOYO-1 (Figure 2 B). Further comparison in the mean fl uorescence intensity indicated that the cellular delivery   The electrostatic attractiveness of cell membrane to polycation/DNA nanocomplex plays a predominant role for its cellular entry. In principle, higher charge of positive nanocomplex benefi ts the intracellular delivery though it usually gives rise to the damages to cells. [ 18 ] As described above, AuNP attachment onto the nanocomplex reduced the zeta potential from 27 mV down to a neutral-close level of 5 mV. When taking into account the sharp decline of zeta potential, the observed enhancement of delivery effi ciency resulting from AuNP attachment seems confusing. There ought to exist other factors that affect the intracellular transport of carried DNA and thereby complement the sacrifi ce caused by the reduced surface charge. Recent studies have identifi ed that the rough surface patched by glycoprotein spikes constitutes a "nanoecology" topography for many enveloped viruses, which is thought to contribute positively to their cell entry. [ 19 ] This speculation can fi nd the evidence that the silica nanoparticle displayed a substantial promotion of cellular entry effi ciency after calcinating relatively smaller ones onto its surface. [ 11 ] Likewise, the easier cellular internalization of the AuNP-attached nanocomplex might associate with the resulting roughness of nanocomplex topography though the defi nite mechanism may be more complicated.
Being aware that the surfi cial AuNP clustering can facilitate the cell uptake of PEI800/pGL-3 nanocomplex, it is interesting to make clear the resulting infl uence on the transfection performance. The in vitro transfection was therefore studied by luciferase assay in the 10% serum-containing medium. As expected, AuNP attachment made a positive impact on the transfection so that the protein expression level was dramatically enhanced in all of the four transfected cell lines at the mass ratio of AuNP versus nanocomplex of 7:1 ( Figure S2B, Supporting Information). As far as the promotion effect of pGL-3 protein expression was concerned, the best outcome was detected in COS7 cells with approximately a 100-fold enhancement. This correlated well with the fi nding that the AuNP-induced enhancement of cell uptake was more distinguished in COS7 cells, validating the feasibility of improving the transfection by virtue of promoting the intracellular plasmid transport. Enhanced green fl uorescent protein plasmid (pEGFP-C1) was also used as the reporter gene for the transfection tests and the transfected cells were observed by CLSM. Compared with the controls using PEI1800/pEGFP-C1 nanocomplex, inside the cells transfected with PEI1800/pEGFP-C1/AuNP appeared signifi cantly stronger green fl uorescence regardless of the cell types ( Figure 3 A), which agreed well with the quantitative lucerifase assay using pGL-3 reporter gene.
The relationship between the transfection effi cacy and the introduced AuNP dosage has been explored ( Figure S2A, Supporting Information). It was shown that as elevating the weight ratio (AuNP vs nanocomplex), the transfection effi cacy took on a gradual increment followed by a plateau starting from the ratio at around 6:1. This fi nding was consistent with the obtained DLS data in terms of the zeta potential. It was found that the variation profi le of the zeta potential also involved two stages, where the zeta potential intended to be insignifi cantly changed after the ratio reached the same ratio ( Figure S3B, Supporting Information). Theoretically, the topography of PEI800/pGL-3/ AuNP nanoparticles would be slightly affected when the surficial AuNP attachment is close to saturation. Thereafter, further increasing AuNP amount would make few infl uences on the surface potential and the transfection effi ciency. Based on the obtained results, therefore, a relatively higher ratio of 7:1 was exploited throughout this study to ensure the complete coverage of the complex surface with AuNPs. On the other hand, DLS data showed that the diameter of the hybrid NPs kept slowly growing along with the ratio increase. The continuous increase www.MaterialsViews.com www.advancedscience.com Adv. Sci. 2015, 2, 1500108   Figure 3. A) Microphotographs of enhanced green fl uorescent protein expression mediated by PEI1800/pEGFP-C1 and PEI1800/pEGFP-C1/AuNP. The concentration of pEGFP-C1 was fi xed at 1 µg mL −1 . The w/w ratio of AuNP versus DNA/vector nanocomplex was fi xed at 7:1. B) CLSM images of the HeLa cells after the incubation with free AuNPs or PEI1800/DNA/Au nanoparticles. The weight ratio of AuNP versus nanocomplex was 7:1. C) Photothermal effects in HeLa cells. 1 min irradiation was applied to cells after the coincubation with samples at various w/w ratios of AuNP versus DNA/vector nanocomplex for 20 h. The laser power was 2 W cm −2 and the concentration of HeLa cells was 5.0 × 10 4 cells per well; the N/P ratio of PEI1800 versus pGL-3 was 20, the pGL-3 concentration was fi xed at 1 µg mL −1 .
of the particle diameter from 210 to 270 nm after the ratio reached 6:1 may possibly associate with the minor aggregation at a higher AuNP dosage ( Figure S3A, Supporting Information). However, such a slight change in the diameter appeared to insignifi cantly affect the transfection effi cacy ( Figure S2, Supporting Information).
A tetrazolium dye (MTT) based assay was carried out to study the infl uence of AuNP attachment on the cell toxicity of PEI1800/pGL-3 nanocomplex. As shown in Figure S4 (Supporting Information), the cell proliferation was nearly not affected by AuNP attachment. The relative cell viability of AuNP-free and -present nanocomplexes always remained above 90% in all of the four cell lines, suggesting the low cytotoxicity originating from the good biocompatibility of PEI1800 and AuNPs.
Rapidly proliferating tumor cells are more sensitive to heat shock than slowly proliferating cells through the mechanisms such as mitotic delay, cell cycle arrest, and plasma membrane damage. [ 20 ] Several studies have shown that the performance of NIR photothermal therapy is sensitive to the size and aggregation state of the gold nanoparticles. Au clusters can elicit coupled surface plasmons and strong absorption at NIR wavelengths with effi cient heat conversion, in contrast to the plasmon resonance of free spherical AuNPs within visible range. [ 21,22 ] It has been identifi ed that gold nanoshells, which was arranged to cover inorganic silica/Fe3O4 nanocore, can offer strengthened NIR absorption, effi ciently convert the light into localized heat, and thus effectively kill cancer cells. [ 23,24 ] Also, AuNP aggregates self-assembled on cell surface were found to offer an increased photothermal effi cacy under NIR irradiation. [ 25 ] In our hybrid nanosystem, "gold nanoclusters" can be formed, where many AuNPs are clustered together on the nanocomplex surface, creating opportunities for photothermal therapy with enhanced NIR absorption effi ciency. To verify this, cancerous HeLa cells were used as the typical model and the cell viability was monitored upon the exposure to CW diode laser irradiation at 808 nm, [ 26 ] which is in the optimal penetration window of most biological tissues. [ 27 ] All the experiments were conducted under the identical conditions and the cells were coincubated with the samples for 20 h prior to the irradiation taken with a very short period of 1 min. It was found that the irradiation treatment imparted negligible infl uences on the cell proliferation when the cells were exposed to either free AuNP or parent PEI1800/pGL-3 nanocomplex, indicating the incapability of them for photothermal induction under the condition (Figure 3 C). Oppositely, the exposure to irradiation led to a robust cell-killing effect when the cells were treated with PEI1800/pGL-3/AuNP, revealing that the success of NIR phototherapy totally relied on the achievement of clustering AuNP on PEI1800/pGL-3 nanocomplex. The weight ratios above 6:1 (AuNP vs DNA/vector nanocomplex) were suffi cient to kill almost all of the HeLa cells. The result correlated fairly well with the aforementioned relevance of the transfection efficiency and zeta potential with the weight ratio, corroborating the surface-saturation hypothesis from another perspective. Furthermore, it was suggested that the majority of the AuNPs in the presence of PEI1800/pGL-3 nanocomplex were actually in a sharply different state from free AuNPs, as can be further demonstrated by CLSM images (Figure 3 B). As shown, inside the cells exposed to PEI1800/pGL-3/AuNP ( W AuNP : W complex = 7:1) appeared lots of dark spots representing the clustered AuNPs that cannot be detectable at all for the treatment using free AuNP. In addition, the comparison between TEM images revealed that a portion of AuNPs in the PEI1800/pGL-3/AuNP system still remained in the free state. To a certain extent, this fi nding implied that the weight ratio of 7:1 used in this study can guarantee the saturated coverage of the nanocomplex with AuNPs.
In summary, a virus-surface-mimicking nanotechnology was herein proposed to develop nonviral gene delivery nanosystem by electrostatically coating DNA-entrapped polymeric nanoparticles with AuNPs. The resulting rough periphery composed of AuNP nanoclusters led to the signifi cantly enhanced cellular uptake of DNA payload and consequently up to 100-fold promotion of transfection effi cacy as compared with the AuNP-absent counterpart. Attractively, this hybridization approach allows the facile import of hyperthermia-induction capability into the traditional gene delivery nanosystem, as demonstrated herein by the outstanding potency for the NIR photothermal therapy in cancerous HeLa cells. It is noted that this distinguished advantage highly relied on the success to form gold "nanoclusters" on the hybrid nanosystem, whereas both the parent nanocomplex and the AuNP separately failed under the identical NIR treatments. Further interests toward this nanotechnology would arise from its expandability to other inorganic NPs for the establishment of multipurpose genebased nanoplatforms, e.g., by using iron oxide NPs to combine gene therapy with the magnetic-induced targeting and thermal therapy.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.