Antiviral Activity of Gold/Copper Sulfide Core/Shell Nanoparticles against Human Norovirus Virus-Like Particles

Human norovirus is a leading cause of acute gastroenteritis worldwide in a plethora of residential and commercial settings, including restaurants, schools, and hospitals. Methods for easily detecting the virus and for treating and preventing infection are critical to stopping norovirus outbreaks, and inactivation via nanoparticles (NPs) is a more universal and attractive alternative to other physical and chemical approaches. Using norovirus GI.1 (Norwalk) virus-like particles (VLPs) as a model viral system, this study characterized the antiviral activity of Au/CuS core/shell nanoparticles (NPs) against GI.1 VLPs for the rapid inactivation of HuNoV. Inactivation of VLPs (GI.1) by Au/CuS NPs evaluated using an absorbance-based ELISA indicated that treatment with 0.083 μM NPs for 10 min inactivated ~50% VLPs in a 0.37 μg/ml VLP solution and 0.83 μM NPs for 10 min completely inactivated the VLPs. Increasing nanoparticle concentration and/or VLP-NP contact time significantly increased the virucidal efficacy of Au/CuS NPs. Changes to the VLP particle morphology, size, and capsid protein were characterized using dynamic light scattering, transmission electron microscopy, and Western blot analysis. The strategy reported here provides the first reported proof-of-concept Au/CuS NPs-based virucide for rapidly inactivating human norovirus.


Introduction
Human norovirus (HuNoV) is a leading food and waterborne pathogen that causes nonbacterial, acute gastroenteritis outbreaks worldwide [1][2][3], accounting for more than 21 million illnesses, and contributing to about 70,000 hospitalizations and at least 570 deaths in the United States each year (Centers for Disease Control and Prevention, 2013). Noroviruses are singlestranded RNA, non-enveloped viruses in the Calicivirdae family. They are classified into five genogroups (GI to GV) and further subclassified into genotypes and genetic clusters based on their capsid sequence [1]. Their genetic diversity, low (18 particles or less) infectious dose [4], inactivation of VLPs (GI.1) by Au/CuS NPs was evaluated using an absorbance-based ELISA. Both nanoparticle concentration and VLP-NP contact time were investigated as possible variables that affect the virucidal efficacy of Au/CuS NPs. Changes to the VLP particle morphology, size, and capsid protein were characterized using dynamic light scattering, transmission electron microscopy, and Western blot analysis. The strategy reported here provides the first reported proof-of-concept Au/CuS NPs-based inactivation approach for inactivating human norovirus.
Au/CuS Core/Shell Nanoparticles Au/CuS core/shell NPs were acquired from Professor Wei Chen's laboratory at the University of Texas at Arlington. Detailed reaction conditions and procedures have been reported previously [50], but, briefly, the NPs were synthesized using a two-step method by first growing gold NPs as cores using the seeded growth method and then coating these cores with CuS nanoshell. The core/shell structure was confirmed by high resolution transmission electron microscope (HRTEM) imaging of the Au and CuS lattice planes in the core and the shell, respectively, and by spectrophotometric observation of the characteristic absorption peaks for Au and CuS at 531 and 981 nm, respectively [50]. The synthesized Au/CuS NPs were 2-5 nm in diameter with an initial concentration of 83 μM.

Au/CuS NPs Treatment to GI.1 VLPs
For treatment with NPs, aliquots (17.5 μL) of purified GI.1 VLP suspensions at various concentrations (to reach the final VLP concentrations at 0.37, 3.7, and 5.6 μg/mL) were mixed with 17.5 μL of various Au/CuS NP concentrations ranging from 0.083 to 20.75 μM (to reach the final concentrations ranging from 0.0083 to 2.075 μM) in 1.5 mL microcentrifuge tubes. All solutions were brought to 175.0 μL using 0.01M PBS and continuously agitated at 30-32 RPM using an end-over-end rotator (Dynal Biotech, Inc.; Lake Success, NY) for various treatment times ranging from 10 min to 4 h. Untreated (no NPs) solutions were prepared identically to the VLPs + NPs solutions and used as controls for each treatment time. After treatment, all solutions were centrifuged at 7500 RPM (5283xg) for 5 min using an Eppendorf microcentrifuge (Hamburg, Germany) to separate the NPs and VLPs.

Evaluation of NPs' Antiviral Activity
The virucidal efficacy of the Au/CuS NPs was evaluated using an ELISA method previously reported by our group [51]. Briefly, 50.0 μL of untreated or treated VLP solution was dispersed into the wells of a medium-binding 96-well polystyrene plate (Costar™ 3591; Corning Incorporated, Corning, NY) and incubated at room temperature for 1 h. 50.0 μL of 0.01M PBS was used as a blank. Each well was washed with 0.01 M PBS and blocked with 100.0 μL of Super-Block T20 (PBS) Blocking Buffer (Thermo Scientific). The wells were washed with PBS and sequentially treated with 0.2 μg/mL mAb 3901 anti-GI.1 VLP and 0.1 μg/mL HRP-labeled goat anti-mouse IgG antibody solutions for 1 h. The plate was washed twice with 100.0 μL aliquots of 0.01 M PBS + 0.05% Tween™ 20 between steps. Following the final washing step, each well was reacted with 100 μL of TMB (3,3',5,5'-tetramethylbenzidine) Peroxidase Substrate Microwell Substrate System (KPL, Gaithersburg, MD) for 10 min, filled with 50.0 μL of Stop Solution (KPL, Gaithersburg, MD), and the absorbance of each well was read at 450 nm using a Spectra-Max M5 plate reader (Molecular Devices, Sunnyvale, CA). Reduced absorbance in NP-treated samples (compared to an untreated control) indicated reduced concentration of intact VLPs due to damage to the capsid surface proteins and associated VLP inactivation.

Characterization of Particle Interactions During Treatment
Capsid size was characterized using DLS to determine how the Au/CuS nanoparticles interacted with the VLPs during treatment. 100.0 μL of each NP-treated and untreated VLP solution was analyzed using a Zetasizer Nano ZSP (Malvern Instruments, Westborough, Massachusetts) and disposable BRAND 1 microcuvettes (Sigma-Aldrich). Suspensions were measured six times each using 10-15 readings per measurement, and the resulting data were averaged to obtain a mean size distribution profile for each solution. Comparing the averaged intensity of the scattered light allows for efficient comparison of VLP dissociation under different conditions [52].

Transmission electron microscopic (TEM) imaging
TEM were acquired using a FEI Technai G 2 Twin TEM (Hillsboro, OR) at the Shared Materials Instrumentation Facility (SMIF) at Duke University. To prepare VLPs samples for TEM, a drop of NP-treated or untreated VLP solution was placed on formvar/carbon TEM grids (Electron Microscopy Sciences; Hatfield, PA) for 10 min. All grids were wicked (to remove excess fluid) using filter paper and then stained with 2% uranyl acetate prior to imaging.

Statistical analysis
Statistical analysis was performed using the general linear model (GLM) procedure of the SAS System 9.2 (SAS Institute Inc., Cary, NC, USA). P < 0.05 was considered as significant different.

Inactivation Effect of Au/CuS NP Treatment on VLPs
We first examined the effect of NPs treatment on VLPs using the ELISA method in which the binding capacity of VLPs to the monoclonal anti-GI.1 VLP antibody (mAb 3901) was evaluated. Fig 1 shows the absorbance signal reductions after VLPs (at two concentration levels: 0.37 and 3.7 μg/mL) were treated with Au/CuS NPs at final concentrations ranging from 0.0083 μM to 1.66 μM for 10 min. For the 0.37 μg/mL VLPs, compared to the untreated control, the inactivation effect was apparent at the treatment with 0.083 μM Au/CuS NPs. Also, there was a marked increase in antiviral activity between 0.083 μM and 0.415 μM, and the VLPs appeared to be completely inactivated at treatments with 0.83 μM and higher NPs. For the denser VLPs at 3.7 μg/mL, treatment with higher concentration of NPs (at 0.415 μM) exhibited a marked antiviral activity, but there were reportable absorbance values at the treatments across the entire tested Au/CuS NP concentration range. This indicates that only partial inactivation was achieved even at high Au/CuS NPs concentrations (up to 1.66 μM).
It turned out that the treatment time had a strong effect on VLP inactivation. The results suggested the NP treated VLPs were morphologically and/or antigenically different from the untreated VLP capsids in that the binding capacity to the detection antibody (mAb 3901) was reduced remarkably. The reduced binding capacity could be due to two possible reasons: NPs might bind to the VLP capsids that may serve to block the binding of mAb3901 to VLPs, or the binding of NPs to VLPs may further cause protein degradation thus the loss/damage of epitopes for mAb 3901 binding.

Effect of NP Treatment on VLP Capsid Protein
We then performed Western blotting to examine protein degradation in VLPs upon Au/CuS NP treatment. Fig 3A shows the immunoblot for VLPs treated with different concentrations of Au/CuS NPs, along with controls. Total proteins were separated by SDS-PAGE and subjected to Western blotting using mAb 3901 primary and fluorescent secondary antibodies. It is known that mAb 3901 can bind to either the full-length (58K) capsid protein or a 32K protein fragment in the P domain [24,53]. It recognizes a continuous epitope on the C-terminal of the capsid protein, as it was able to bind to the 58K capsid protein and the 32K protein product in Western blot even when the proteins were denatured by boiling prior to analysis [53]. The Western blot here shows that all samples presented a band near 32K, but the relative intensities of the bands obviously varied for the samples treated with different concentrations of NPs. Taking the band intensity of the VLP samples without NP treatment as a base value (taken as 1), the normalized band intensities of the samples treated with different concentrations of NPs are shown in Fig 3B. The results clearly showed that the amount of this 32K protein product reduced significantly as the VLPs were treated with NPs. The treatment with the lowest concentration of NPs (0.83 μM) in the test caused~86% reduction of this protein, and as the NPs concentration increased to 10 to 20 times higher, it reduced even more and reached a maximum 92-95% reduction. The reduced amount of this protein after NPs treatments suggested that Au/CuS NPs treatment most likely cause degradation (break-down) of this 32K P domain protein.   It is also known that mAb 3901 also recognizes a domain between amino acid 453 and amino acid 495, and the lower band (~22K band) in the Western blot is likely a fragment that contains this sequence. As shown in Fig 3C, this band was less pronounced in the control samples, but was significantly increased in the NPs-treated VLP samples. In fact, normalizing the band intensities relative to the untreated control samples showed that this protein fragment actually increased 2.7-to 3.1-fold in the NPs-treated samples. Again this observation suggested it is likely that the NP treatments caused the 32K P domain protein to break down into smaller fragments. The~22K product was possibly one of the fragments and contained the domain (aa 453-495) that was recognized by mAb3901. It is not surprising that NPs treatment caused VLP capsid protein degradation, as several other inactivation methods treatments have been reported to cause VLP capsid protein degradation, such as Gamma radiation [18], high pressure treatment [17]. And especially, capsid protein degradation in human norovirus was observed upon contact with copper alloy surface [54] which is closely relevant to this study. At this stage, however, it is unclear about the detailed mechanisms on how Au/CuS NPs can cause the capsid protein to break down, and further studies are necessary to elucidate the detailed mechanisms.

NPs Treatment Damages Virus Particles
We further investigated the particle size profiles in VLP suspensions before and post-NPs treatment using dynamic light scattering (DLS). Fig 4 shows the mean particle diameter profiles of VLP suspension, Au/CuS NPs suspension, and the VLP suspension after NP treatment. As expected, for the VLP suspensions, the peak between 10 to 100 nm represented the VLP particles in the suspension, as it is known that the diameter of an assembled norovirus VLP is~38 nm [53], but smaller and larger particles with 20 to 90 nm diameter were also observed [55]. For the Au/CuS NP suspensions, the peak between 1 and 10 nm indicated the profile of the NPs diameters, which was very close to the previously reported size of 2-5 nm (as determined by SEM/TEM). For VLPs + Au/CuS NPs suspensions, it is obvious that the VLP peak disappeared after 10 min of mixing, suggesting that the intact VLPs may have broken into smaller fragments. Tests on VLP suspensions treated with NPs at other concentrations (0.83 μM and 1.245 μM) showed similar results where the peak between 10 and 100 nm before treatment shifted to <10 nm after NPs treatments (S1 Fig). Damage to VLPs upon Au/CuS NP treatment was confirmed using TEM ( Fig 4B). As shown in the images, the untreated VLPs were intact and readily identifiable with well-defined, regular round shape. However, these features were less distinct after treatment with 0.083 μM Au/CuS NPs. The treatment with the 0.83 μM NPs appeared to break the VLPs' capsids into fragments and no intact VLPs were observed. The imaging results confirmed that the VLP capsid underwent physical degradation and eventual rupture as the nanoparticle concentration increased from 0.083 μM to 0.83 μM (Fig 4B). This cumulative loss of capsid structural integrity is consistent with the particle size profile changes observed in the DLS spectra.

Possible Mechanisms of Au/CuS NP Inactivating VLPs
Based on the reported mechanisms of how NPs inactivate bacteria, direct contact and damage to the cell membrane is a common mechanism for many different NPs [56][57][58]. Other possible routes include suppression of energy metabolism [59], inhibition of enzyme activity and induced oxidative stress [57], increased membrane permeability [60][61][62], and physical piercing [63]. Although these mechanisms might not all be applicable to how Au/CuS NPs inactivate VLPs, it is likely that some of these mechanisms are involved. The above observations suggested that the Au/CuS NPs may inactivate the VLPs by direct contact/binding to the VLPs and further physically damaging the capsid.
Considering the core/shell structure of the Au/CuS nanoparticles, the CuS shell is the component that most likely interacts with the VLP capsid. Little is known about the antimicrobial activity of pure CuS nanoparticles based on published literature. However, another member of the Cu compound family, the CuO NP, has recently been reported to show antimicrobial activity to several types of microorganisms [64][65][66]. Based on the action of CuO NPs and other metal oxide NPs (such as ZnO), the release of soluble metal ions (such as Cu 2+ , Zn 2+ ) from the NPs largely influenced their toxicity [64]. Our previous study on Au/CuS NPs inactivation of B. anthracis spores and cells showed that Au/CuS NPs bind to and damage the cell membrane, creating an osmotic imbalance, efflux of cytoplasmic content, and associated cell rupture and death [49]. More relevant studies using copper alloys as antiviral surfaces for murine norovirus -an HuNoV surrogate-showed the virus was destroyed within 60 min [67], with the largest difference in inactivation rate occurring within the first 30 min of contact [68]. These studies reported that the mechanism of copper inactivation of noroviruses involves both degradation of the RNA and destruction of the capsid. A more recent study also reported that exposure to copper alloys destructed the capsid and genome of GII.4 human norovirus [54]. These reports suggested that the Cu component may be a key factor controlling the antiviral activity of Au/ CuS NPs to VLPs. And our observations on VLPs' capsid protein degradation and the loss of capsid integrity by Au/CuS NPs treatment are consistent with the observations reported by these relevant studies [54,[67][68][69].
It is also possible that an intact NP can diffuse across the cell membrane or virus capsid, or that Cu 2+ solute from the NPs can enter cells through the transport and ion/voltage-gated channels [70]. While NPs themselves can interact with oxidative organelles or redox active proteins to induce reactive oxygen species (ROS) in cells, Cu 2+ produced by the NPs can also induce ROS by various chemical reactions, and ROS can break DNA strands and alter gene expression [64]. Another possible mechanism is that Cu 2+ can chelate with biomolecules or dislodge the metal ions in some metalloproteins, leading to dysfunctional proteins and further cell inactivation [64]. These mechanisms are likely applicable to how Au/CuS NPs interact with VLPs, leading to capsid protein degradation and breakdown. However, as the study of Au/CuS NPs' antiviral activity is in a very early stage, its detailed mechanisms are not fully understood. Further studies using different viruses and experimental conditions, systematic comparison with pure Au NPs, CuS NPs, and other similar metal NPs, are necessary to fully understand the virucidal properties of Au/CuS NPs.
This study only demonstrated the proof-of-concept that Au/CuS NPs exhibited antiviral activity to norovirus VLPs, and much more work is still needed to apply the concept to a practical approach. Nevertheless, the observations here present both challenges and opportunities for further investigations. The challenges would be to elucidate the detailed mechanisms about Au/CuS NPs' antiviral activity to VLPs, and to investigate its antiviral activity to norovirus surrogates and native human noroviruses. Since using VLPs as the test model would not be possible to study the effect of Au/CuS NPs on the infectivity of noroviruses, further studies on this aspect must be performed using surrogates and/or human noroviruses. On the other hand, the opportunities are such that there is still much room for improvements from the proof-of-concept results in this study if the antiviral activity is confirmed using norovirus surrogates and human noroviruses. There is also opportunity to develop approaches for effective incorporation/utilization of these NPs in practical antiviral agents would be needed. Potential applications of such NPs-incorporated antiviral agents would be disinfectants/sanitizers for decontaminating norovirus contaminated surfaces, especially in hospital or clinical settings, since contaminated surfaces have been reported as a secondary outbreak in many reported norovirus outbreaks because of inadequate disinfection [54,71]. Rapid and effective antiviral action of NPs-based disinfectants would be useful in preventing the spread of infection by reducing secondary transfer from contaminated surfaces in these settings.

Conclusions
The current lack of rapid, point-of-care inactivation strategies hinders progress in controlling frequent and widespread norovirus outbreaks. Because the most promising virucides directly bind to the viral capsid [43], inactivation via nanoparticles (NPs) is a more universal alternative to other physical and chemical strategies that exhibit variable efficacy. Using GI.1 VLPs as a model viral system, this study provides proof of concept that Au/CuS core/shell NPs can rapidly inactivate human norovirus. Immunoblotting, dynamic light scattering, and TEM results provided evidence that capsid protein degradation and capsid damage appeared to be the mechanisms associated with inactivation and have a direct dependence on both NPs concentration and treatment time. Nevertheless, this study demonstrated that Au/CuS NPs are promising antivirals, although further studies using HuNoV-rich fecal extracts will be beneficial to confirming the efficacy of Au/CuS NPs against norovirus.