Investigating Silver Nanoparticles and Resiquimod as a Local Melanoma Treatment

Abstract

Over the last decade, the potential for silver nanoparticles (AgNP) to be used as an anti-melanoma agent has been supported by both in vitro and in vivo evidence. However, an undesirably high concentration of AgNP is often required to achieve an antitumor effect. Therefore a combination treatment that can maintain or improve antitumor efficacy (with lower amounts of AgNP) while also reducing off-target effects is sought. In this study, the combination of AgNP and resiquimod (RSQ: a Toll-like receptor agonist) was investigated and shown to significantly prolong the survival of melanoma-challenged mice, when added sequentially. Results from toxicity studies showed that the treatment was non-toxic in mice. Immune cell depletion studies suggested the possible involvement of CD8⁺ T cells in the antitumor response observed in the AgNP + RSQ (sequential) treatment. NanoString was also employed to further understand the mechanism underlying the increase in the treatment efficacy of AgNP + RSQ (sequential). Results showed significant changes in gene expression involving apoptosis and immune stimulation pathways when compared to the naïve group. In conclusion, the combination of AgNP and RSQ is a new combination worthy of further investigation in the context of melanoma treatment.

Graphical abstract

Introduction

Melanoma persists as the deadliest form of skin cancer despite constant research into cures and the availability of a wide range of treatment strategies [1, 2]. Surgical excision can be considered the most effective local treatment for melanoma patients; however, there are cases where multiple surgeries cannot stop the disease from reoccurring [3, 4]. Chemotherapy is another common approach particularly when the disease has progressed to being invasive or metastatic; nevertheless, melanoma can develop resistance to these drugs over time [5–7]. Checkpoint blockade is a relatively recent immune-based therapy displaying objective responses in patients with advanced melanoma of 10 – 58% [8, 9]; nonetheless, the non-inflamed melanoma tumor microenvironment (TME) [10] and the immunosuppressive mechanisms of melanoma cells and other cells (e.g. tumor-associated macrophages and regulatory T cells) within the TME can limit the treatment efficacy of checkpoint blockade. Combination therapies can improve therapeutic outcomes for melanoma patients compared to monotherapy. For example, the combination of the checkpoint inhibitors, nivolumab (anti-PD-1) and ipilimumab (anti-CTLA-4) was shown to have higher melanoma treatment efficacy than when each treatment was used alone [11]. Other combinations demonstrating improved therapeutic outcomes compared to monotherapy include dabrafenib + trametinib [12], imiquimod + IL-12 [13], and temozolomide + IFNα2 [14]. In this study, the effect of silver nanoparticles (AgNP) and resiquimod (RSQ) on melanoma progression was investigated.

RSQ is an immunostimulatory molecule that binds to Toll-like receptor (TLR) −7 and −8, and is expressed primarily by cells of the innate immune system such as macrophages and dendritic cells (DC); triggering the release of a suite of cytokines, including IL-12 and IFN-ɤ [15]. It has been shown in both preclinical and clinical studies that the anti-melanoma activity of RSQ is associated with its ability to enhance Th1-biased immune responses [16–21]. These studies also suggest that RSQ is more efficient at promoting antitumor activity when combined with a treatment that induces at least some degree of tumor cell death and thereby provides a source of antigens for the immune system to recognize [20, 22].

AgNP have many biomedical applications, including acting as carriers for antitumor agents [23] and mediating photodynamic therapy [24]. The ability of AgNP to directly kill melanoma cells was first reported in 2013 and studies have suggested that AgNP can induce melanoma cell death through apoptosis in a dose-dependent manner in vitro [25, 26]. A dose-dependent effect has also been shown in vivo where AgNP were administered subcutaneously to melanoma-challenged mice; and high concentrations of AgNP (12 mg/kg) were also shown to be safe [25]. However, many studies into AgNP toxicity suggests AgNP at high concentration (up to 1 g/kg) can mediate dose-dependent toxicities [27–29]. Hence, low concentrations of AgNP may be more acceptable when used as a therapeutic agent in the clinic as this would reduce the chances of undesirable side-effects.

Both AgNP and RSQ, while having anti-melanoma activity when used independently as monotherapies, have been found to possess even higher therapeutic efficacy when each are used in combination with other agents [30, 31]. An example is when AgNP were used in combination with wortmannin (an anti-autophagy agent and PI3 kinase inhibitor) [31]. Results in melanoma-challenged mice showed that the combination can significantly delay melanoma growth as compared to when each treatment was used alone [31]. However, wortmannin is unlikely to translate into the clinic due to its toxicity, and thus more biocompatible drug combinations involving AgNP are sought, such as RSQ (and AgNP) [32]. RSQ, aside from its immune properties noted above, is also considered safe when delivered locally [33]; and when topically applied has shown promising results in clinical trials [34]. Also, the combination of AgNP and RSQ can be delivered topically, and thus represents a less invasive treatment strategy, that may stimulate greater patient compliance [34–37] .

Here, we investigated whether combining AgNP and RSQ could lead to an enhancement in treatment efficacy based on the drugs’ different mechanisms of action. To explain, AgNP, through direct apoptotic-mediated killing of tumor cells, should provide a source of tumor antigens which can be taken up by antigen-presenting cells, such as DC, which upon activation with RSQ, can potentially promote tumor-specific T cell responses. It was shown using a murine melanoma model that the sequential administration of AgNP and RSQ could significantly extended survival compared to untreated mice, and that CD8⁺ T cells may be involved in the anti-tumor activity. Further investigation, using NanoString analysis, also indicated significant changes in the levels of apoptosis and type I immune response related genes, when the combination group (sequential administration of AgNP and RSQ) was compared with the naïve group.

Materials and Methods

Materials

Silver nitrate (AgNO₃), polyvinylpyrrolidone (PVP, Mw = 55,000 g/mol), N-acetylcysteine (NAC), and dimethyl sulfoxide (DMSO) were bought from Sigma-Aldrich (St. Louis, MO). Sodium borohydride (NaBH4) was obtained from Fisher Chemical (Fair Lawn, NJ). RSQ was obtained from Selleckchem (Houston, TX). Succinate buffer (0.2 M; pH 5.5) was procured from Alfa Aesar (Ward Hill, MA). All materials were used as purchased unless stated otherwise.

B16.F10 (a murine melanoma cell line) and A375 (a human melanoma cell line) were bought from the American Type Culture Collection (Manassas, VA). Dulbecco’s Modified Eagle Medium (DMEM medium), 1 M HEPES, 100x GlutaMax^™, 100 mM sodium pyruvate, 100x MEM non-essential amino acid solution (MEM NEAA), 100x penicillin-streptomycin, and 1x Dulbecco’s phosphate buffered saline (DPBS) were acquired from Gibco (Grand Island, NY). Fetal bovine serum (FBS) was purchased from Atlanta Biological (Flowery Branch, GA). Gentamycin sulfate was acquired from IBI Scientific (Dubuque, IA). The DMEM medium used to culture B16.F10 cells contained 10% v/v FBS, 10 mM HEPES, 2 mM GlutaMax™, 1 mM sodium pyruvate, and 50 ng/ml gentamycin sulfate. The DMEM medium used to culture A375 cells contained 10% v/v FBS, 0.1 mM (per each amino acid) MEM NEAA and 100 μg/ml penicillin-streptomycin. All cells were incubated at 37°C and 5% CO₂ unless stated otherwise. Medium containing all supplements (including FBS) is referred to as complete medium.

AgNP synthesis and characterization

AgNP were synthesized using a chemical reduction method. PVP (560 mg in 20 ml Nanopure water (Barnstead International, IA) and AgNO₃ (38.2 mg in 7.5 ml Nanopure water) solutions were added together and stirred at the highest speed on a stir plate (IKA^®-Werke GmbH & Co. KG, Germany) in the dark for 1 hour. NaBH₄ solution (60 mg in 6.5 ml Nanopure water) was rapidly injected into the mixture using a 5 ml syringe and 26G needle. The suspension was stirred for another 1 hour in the dark and transferred to a dialysis bag (M_w cutoff = 3500 Da). The dialysis process was performed against Nanopure water in the dark for 3 days, changing the water twice daily. The AgNP solution was lyophilized using a lyophilizer (Labconco, MO) and resuspended in an appropriate medium when used. All the solutions used were freshly prepared prior to the synthesis process. The synthesis process is depicted in Figure 1a.

Figure 1:

AgNP synthesis and characterization. (a) Schematic showing the AgNP synthesis process (more details can be found in the Materials and Methods). (b) Photo of different solutions used in AgNP synthesis process (I-III; I, II, and III represent AgNO₃, NaBH_4> and PVP solutions, respectively) and all the reagents combined (IV). The distinct difference in solution color in IV indicated the presence of AgNP. (c) The absorption spectra of different solutions used in the AgNP synthesis process (I-III; I, II, and III represent AgNO₃, NaBH₄, and PVP solutions, respectively), and all the reagents combined (IV). Note that for solution (IV), the absorption spectrum was measured at the end of the dialysis process. (d) The morphology of AgNP as measured by TEM. (e) The size distribution of AgNP as measured by ImageJ software (mean diameter: 10.60 nm).

AgNP were resuspended in Nanopure water for the characterization process. A cuphorn sonicator (Qsonica sonicators, Newtown, CT) was used at an amplitude of 75 for 2 minutes to resuspend the AgNP. The absorption spectra were measured using the scanning mode at a wavelength range of 300–700 nm using UV-visible spectroscopy (Spectra Max^® plus 384 Microplate Spectrophotometer, Molecular Devices, CA). The morphology of the AgNP was obtained using a transmission electron microscope (TEM, JEOL USA, Inc. MA) and the size of the AgNP was determined by analysis of TEM images using ImageJ software. The Zetasizer Nano ZS particle analyzer (Malvern Instrument Ltd., Westborough, MA) was also used to measure the hydrodynamic diameter and the zeta-potential of the AgNP. The silver content in AgNP was measured using inductive coupling plasma mass spectroscopy (ICP-MS, Agilent Technologies, Santa Clara, CA). The sample preparation protocol for this process can be found in the supplementary section. The AgNP used in the in vitro and in vivo studies were resuspended in complete media and 0.2 M succinate buffer, respectively.

Anti-proliferative effects of AgNP on melanoma cell lines (B16.F10 andA375)

B16.F10 and A375 cells in complete media were seeded into 48-well plates at a density of 5,000 cells/200 μl/well and incubated overnight. AgNP were added at varying concentrations (10–500 μg/ml) to the indicated wells. After 4 or 24 hours incubation, the supernatant was discarded, the cells were washed twice with DPBS, and the cell viability was assessed using the PrestoBlue assay (ThermoFisher Scientific, Eugene, OR) according to manufacturer’s instructions. The fluorescence signal was measured at the excitation/emission of 560/590 nm using a Spectra Max M5 fluorescence microplate reader (Molecular Devices, CA). The ratio between the signal read from the treated and untreated cells (100% viable) was calculated as the relative viability and expressed as a percentage. The background readout (from the wells containing only complete media and PrestoBlue reagent (no cells)) was subtracted from the sample readouts prior to analysis. AgNP has been reported to rapidly induce apoptosis and peak at 6 hours of incubation [25]. Thus, to observe the effects of AgNP on melanoma cell viability, both 4 and 24 hour incubations were included in the study.

Reactive Oxygen Species (ROS) production

Reactive oxygen species (ROS) detection assay kit (Biovision, CA) and N-acetylcysteine (NAC) were used to evaluate the ability of AgNP to induce ROS production in melanoma cells. The assay was performed according to the manufacturer’s protocol. In brief, B16.F10 cells in complete media were seeded into the wells of a 96-black well clear bottom tray (Thermofisher Scientific, Eugene, OR) at 5000 cells/100 μl/well and incubated overnight to allow for cell attachment. The supernatant was aspirated, and the cells were treated with the ROS label. The cells were then incubated in the dark for 45 minutes, the label removed, and different concentrations of AgNP (4 and 20 μg/ml) were added to the cells. After 6 hours incubation, the fluorescence signal was measured at the excitation/emission wavelengths of 495/529 nm. The kit provided an ROS inducer that was used as a positive control; it was added to the B16.F10 cells without any addition of AgNP 1 hour prior to the analysis process. The percentage of ROS production was calculated using equation 1.

$$PercentageofROSproduction=\frac{Signalreadfromsample-Signalreadfromcontrol}{SignalreadfromROSinducer-Signalreadfromcontrol}$$

Equation 1: ROS inducer represented positive control provided in kit; control represented no AgNP or ROS inducer.

For the NAC assay, B16.F10 cells in complete media were seeded into the wells of 48-well plates at 5,000 cells/200 μl/well and incubated overnight to allow for cell attachment. Then NAC (0.17 mg/ml in complete media) was added to the cells and incubated for 30 minutes; then AgNP was added at increasing concentrations (10–500 μg/ml) to the wells without removing NAC. After 24 hours of incubation, the supernatant was discarded, the cells were washed twice with DPBS, and the PrestoBlue assay was performed to assess the cell viability as per manufacturer’s instructions.

Caspase assay

The caspase assay was performed using a caspase-3, caspase-8, and caspase-9 Multiplex Activity Assay Kit (Fluorometric, ab219915, abcam, MA). The assay was performed according to the manufacturer’s protocol with some modifications. In brief, B16.F10 cells in complete media were seeded into the wells of a 96-well tissue culture tray (5,000 cells/100 μl/well) and incubated overnight to allow for cell attachment. The supernatant was aspirated and then different concentrations of AgNP (4–500 μg/ml) were added and incubated with the cells for 6 and 24 hours (100 μl/well). Caspase substrate was added to the designated well (100 μl/well) without removing the preexisting media in the wells and the plate was incubated at room temperature for 30 minutes protected from light. Note that different caspase substrates were added to different wells. The signal was read using a Spectra Max M5 fluorescence microplate reader (Molecular Devices, CA) at the excitation/emission of 535/620 nm, 490/525 nm, and 370/450 nm for caspase-3, −8, and −9, respectively. To minimize possible interference from AgNP, the signal read from AgNP at different concentrations was subtracted from the wells containing the substrate + AgNP without any cells before the analysis process.

In vivo effect of the combination of AgNP and RSQ on tumor progression and survival of melanoma-challenged mice.

B16.F10 cells (10⁵ cells in 100 μl in complete media without FBS) were injected subcutaneously (s.c.). into the right dorsal flank of C57BL/6j female mice (weight 18–20 g, 6–8 weeks old). The mice were divided randomly into the following groups: naïve, succinate buffer (vehicle), AgNP, RSQ, AgNP+RSQ (both agents delivered concurrently), and AgNP+RSQ (sequential: RSQ delivered 24 hours after AgNP). The treatment started on day 7 post tumor challenged (PTC) and was repeated at 3-day intervals for a total of 7 treatments. Both AgNP (2 mg/kg) and RSQ (25 mg/kg) were resuspended in 0.2 M succinate buffer and administered to the mice in a volume of 100 μl. The administration routes used were peri- and intratumoral when the tumor not measurable and palpable, respectively. The tumor volumes throughout the study were calculated using equation 2. The mice were sacrificed when: the tumor reached the predetermined size (20 mm in width or length or height of 10 mm); there was significant weight loss in mice (more than 20%); or the tumor was ulcerating/bleeding. Mice sera were collected from the submandibular vein on day 21 PTC and sent to the IDEXX company (Westbrook, Maine, USA) to perform toxicity analysis. All in vivo experiments were performed according to the guidelines and regulations approved by the University of Iowa Institutional Animal Care and Use Committee. The study plan described here is depicted in figure 4a. It should also be noted that the mice treated with PVP alone were not included as it had been reported to not affect tumor growth nor have any toxic effects in mice [25].

Figure 4.

Tumor growth and survival study. (a) Schematic showed the challenge and treatment schedule. (b) The individual tumor volume of melanoma-challenged mice receiving different treatments (c) Tumor volumes in mice receiving different treatments over time up until day the last day that all the mice were still alive (i.e 19 post tumor challenge (PTC)). The error bars represented the standard error. (d) Average tumor volume (Day 19 PTC) of mice receiving different treatments. Dunn’s multiple comparisons test showed that there was a significant difference between the group treated with AgNP+RSQ (sequential) compared to the naive treated group (p-value of 0.0396). Error bars shown in this graph represented the standard error. (e) Survival analysis of melanoma-challenged mice receiving different treatments. The p-values for the naive and succinate groups, when compared to the AgNP+RSQ (sequential) group, were 0. 0103 and 0.0187, respectively. More details can be found in Table S5. (n = 9/10 per group). * indicated P-value ≤ 0.05.

$$Tumorvolume\text{=}\frac{\pi }{6}xHeightxWidthxLength$$

Equation 2: The equation used to calculate tumor volume.

Lymphocyte subset depletion study

The mechanism underlying the anti-tumor effect of the combination treatment, AgNP+RSQ (sequential), was investigated using selective immune cell depletion. The experimental procedures, including mouse strain, the tumor challenge protocol, the treatment intervals, and euthanization criteria were similar to that described in the survival section. In this study, the mice were randomly divided into 4 groups: naïve, AgNP+RSQ (treatment); treatment+anti-CD8; treatment+anti-CD4. Antibodies against the indicated lymphocyte subsets were intraperitoneally injected (150 μg/antibody/mice/injection) before and throughout the treatment. The study described here is depicted in figure 6a. It is of note that antibodies used in this study were produced in the Salem’s laboratory. The hybridoma clones used to produce anti-CD4 and anti-CD8 antibodies were GK 1.5 (ATCC), and 2.43 (ATCC), respectively. Blood from mice was collected on day 7 PTC (the day the treatment started) to assess the percent of T cells depleted. The results showed to be 86–93% and 96–99% for CD4+ T cells and CD8+ T cells, respectively.

Figure 6: Lymphocyte subset depletion study.

(a) Schematic showing the challenge and treatment schedule. The numbers on top of the arrows indicate the days the antibodies were intraperitoneally injected. (b) The individual tumor volume of melanoma-challenged mice receiving different antibodies, (c) Tumor volumes in mice receiving indicated lymphocyte subset-specific antibodies over time up until the last day that all the mice were still alive (i.e.16 days post tumor challenge (PTC)). No significant difference was found between different treatment groups. The error bars represented the standard error. (d) Average tumor volumes (Day 16 PTC) of mice receiving different treatments. Error bars shown in this graph represented the standard error. (e) Survival analysis of melanoma-challenged mice receiving different antibodies. More detail on statistical results can be found in Table S8. (n = 9–10/treatment). Color should be used in print.

RNA extraction and NanoString study

The experimental procedures, including mouse strain and tumor challenge were the same as described in the survival and lymphocyte subset depletion study sections (n=6/group). Mice were randomly divided into two groups after tumor challenge: naïve (untreated) and AgNP+RSQ (sequential) treated mice. The euthanization criteria for the mice in the naïve group were as described in a previous section; however, all the remaining mice in the treatment groups were euthanized once all the naïve mice were euthanized. The tumors from all mice were collected and immediately stored in RNAlater^™ Stabilization Solution (Invitrogen, MA) at −80°C until the RNA extraction process was performed.

For the RNA extraction process, the tumors were removed from the RNAlater^™ Stabilization Solution approximately 100 mg per sample and was placed into a 2 ml microcentrifuge tube. TRIzol^™ Reagent (Invitrogen, MA) was added to the sample (1 ml/50–100 mg) and homogenized using a tissue homogenizer (Tissue Tearor^™, BioSpec Products, Inc, Bartlesville, OK) using level 4 for 30 seconds to 1 minute at room temperature. The samples were incubated at room temperature for 5 minutes and then chloroform was added to the sample (200 μl/1 ml TRIzol^™). The samples were then incubated at room temperature for 3 minutes and centrifuged at 12,000 g for 15 minutes. The aqueous phase was transferred to a new RNAse-free microcentrifuge tube and 1 equal volume of 70% ethanol was added to the tube. An RNeasy Mini Kit (Qiagen, Germantown, MD) was used to isolate RNA and the samples were cleaned and concentrated with RNA clean & Concentrator-25 (Zymo Research, CA) according to the manufacturer’s protocols. The samples were sent to the Iowa Institute of Human Genetics, University of Iowa Carver College of Medicine for NanoString analysis using a PanCancer Immune Profiling Panel. Data analysis methodology can be found in the supplementary section. NanoString data were blindly analyzed by a third party.

Statistical Analysis

One-way analysis of variance (ANOVA, Dunnett and Tukey’s multiple comparisons) and the paired t-test were employed for in vitro studies using GraphPad Prism software. Log-rank test with multiple comparisons was utilized for analysis of mouse survival, using SAS software. The tumor volumes at the last day all the mice were still alive were also compared using one way analysis of variance (ANOVA, Dunn’s multiple comparisons). The NanoString data was analyzed using R studio software (RStudio Team (2020). RStudio: Integrated Development for R. RStudio, PBC, Boston, MA URL http://www.rstudio.com/); details of the analysis can be found in the supplementary section. Significant differences were assigned at p-values < 0.05. The error bars were shown as standard deviation unless stated in the figure captions.

Results and discussion

AgNP synthesis and characterization

The use of AgNP as an anti-melanoma agent was first reported in 2013 [26]. However, the physicochemical properties of AgNP can vary depending on size, shape, and capping agent used [20, 38, 39]. PVP-capped AgNP have been reported to have direct dose-dependent anti-melanoma activity in vitro and in vivo [25, 31]. PVP-capped AgNP were also reported to be non-toxic in vivo (when administered to mice subcutaneously (up to 12 mg/kg)), and in vitro (using a keratinocyte cell line, HaCaT; 100 μg/ml for 24 hours) [25, 39]. Hence, PVP-capped AgNP were selected for this research.

In the present study, AgNP were synthesized adopting chemical reduction methods. The change in solution color from clear to yellow/brown (Figure 1b) indicated the presence of AgNP [40, 41]. The appearance of a peak at 395 nm in the UV-absorption spectra (Figure 1c) confirmed the presence of AgNP as it represents the surface plasmon resonance peak, a unique characteristic of metal nanoparticles [42–44]. Based on the absorption spectra, the synthesized AgNP were speculated to be spherical with diameters < 40 nm [45]. This was confirmed by TEM image analysis (Figure 1d) where the mean diameter of the AgNP as measured by Image J software was determined to be 10.60 ± 4.93 nm (Figure 1e). The hydrodynamic diameter and the zeta-potential of the AgNP were 11 nm (PDI = 0.510) and −26.5 mV, respectively (Figure S1). It has been reported that particles with zeta-potential values > +25 mV or < −25 mV have high repulsion forces between the particles [46]. Thus, the AgNP synthesized can be considered stable. The silver content found in AgNP was showed to be 58.86 ± 5.61% of the initial Ag+ added to the AgNP synthesis process as measured by ICP-MS (Table S1).

In vitro effects of AgNP or RSQ on melanoma cells.

In order to investigate the effects of AgNP on the viability of melanoma cells, two cell lines (B16.F10 (mouse-derived) and A375 (human-derived)) were cultured with increasing concentrations of AgNP and then assayed for viability using the PrestoBlue assay. The results (Figure 2) showed that AgNP can significantly reduce the viability of both melanoma cell lines in a dose-dependent manner. These results indicate that AgNP affect both murine and human melanoma cells, and the magnitude of the effect can be cell line dependent [47]. Similar findings with B16.F10 and A375 have been recently reported by another group where they also showed that AgNP reduced the viabilities of two other human melanoma cell lines, SK-MEL-28 and WM35, with WM35 being the most sensitive to AgNP treatment [48]. The primary mechanism by which AgNP mediate their toxicity is believed to be through dissolution of AgNP into Ag⁺ [49, 50]. Ag⁺ induce ROS production and disrupt mitochondrial homeostasis of Ca²⁺ leading to mitochondrial dysfunction [51, 52]. This can result in cell cycle arrest and cell death, as has been observed in HepG2 cells [53]. Incubating B16.F10 cells with increasing concentrations of RSQ (up to 50 μg/ml) revealed no significant effect on cell viability (Supplementary Figure S2) and was only marginally cytotoxic at extremely high concentrations (250 μg/ml). These findings corroborate with in vitro findings (using B16.F10 or Lewis lung carcinoma cells) and clinical findings (in patients with cutaneous T cell lymphoma) reported by others that RSQ does not mediate its toxic effects directly but rather through immune mediated mechanisms [15, 54, 55].

Figure 2:

The effect of AgNP on melanoma cell viability in vitro (as measured using the PrestoBlue assay). B16.F10 and A375 at 4 and 24 hours incubation. The asterisks indicate significant differences between the two groups being compared, using a paired t-test. ** = P-value < 0.01, *** = P-value < 0.001, and **** = P-value ≤ 0.0001.

AgNP can induce melanoma cell death via an apoptosis-related pathway.

AgNP have previously been reported to induce the death of melanoma cells (B16.F10) in vitro via an apoptotic pathway as determined by an annexin V/PI assay; however, further mechanistic evaluations have not been performed [25]. In this study, measurements of caspase-3, −8, and −9, as well as ROS production, were performed as these assays can delineate pathways of apoptotic cell death [56–58]. Of the caspases, activation of caspase-3 can directly lead to apoptotic cell death [59, 60]. However, changes in caspase-8 and −9 levels can also provide information on the type of apoptosis pathway involved as: 1) they are the initiators of extrinsic and intrinsic/mitochondrial related apoptosis, respectively [61–63], and 2) both caspases, especially caspase-9, may be involved in a caspase-3 independent apoptotic pathway (i.e. the TNF-α pathway) [64].

The results (Figure 3a) indicate the presence of the three caspases in the supernatant of the AgNP treated cells suggesting the involvement of both the intrinsic and extrinsic pathways of apoptosis in AgNP-mediated cell death. The fold increases in caspase-3 and −8 were in the range of 2 – 4-fold while for caspase-9 the fold increases were in the range of 4 – 14-fold, suggesting that the intrinsic pathway may be the more dominant pathway through which AgNP mediate cell death. In addition, after 24 hours, caspase-9 was the only caspase tested to be present at significantly increased levels (7 – 11-fold) when cells were incubated with high concentrations of AgNP (100 and 500 μg/ml). Many studies have shown evidence supporting that AgNP use the intrinsic pathway to induce melanoma cell death in vitro (e.g. A549, HCT-29) [65–69]. Caspase-9 has also been reported to induce apoptotic cell death via the TNF-α pathway, an apoptotic pathway that can be independent of caspase-3 activation [64], which may explain the decrease of caspase-3 despite the increase in caspase-9 at 24 hours.

Figure 3: Effect of AgNP on apoptosis and ROS production by B16.F10 melanoma cells.

(a) Changes in caspase-3, −8, and −9 levels released when B16.F10 cells were incubated with indicated concentrations of AgNP for 6 and 24 hours (n = 3/group). (b) The percent of reactive oxygen species (ROS) produced after the cells were incubated with different concentrations of AgNP for 6 hours as measured by the Biovision kit (n = 3/group). The AgNP concentrations used in this assay were low as the B16.F10 cells needed to be viable at the time of measurement. (c) The cell viability of B16.F10 cells after pretreatment with N-acetylcysteine (NAC) followed by treatment with indicated concentrations of Ag⁺ or AgNP for 24 hours as measured by PrestoBlue assay (n = 3/group). * indicated p-value ≤ 0.05, ** indicated p-value ≤ 0.01, *** indicated p-value ≤ 0.001, and **** indicated p-value ≤ 0.0001. More detail on the statistical analysis on the cell viability of B16.F10 cells with and without NAC pretreatment and cocultured with different concentrations of Ag⁺ and AgNP can be found in Tables S2 and S3.

AgNP have been reported to induce ROS production in melanoma cells (B16.F10) in vitro [25], which in turn can lead to apoptosis. Thus, in this study, we investigated the ability of AgNP to induce ROS production in B16.F10 cells. The results showed that ROS production was induced by AgNP in a dose-dependent manner (Figures 3b and 3c, Tables S2 and S3). The ROS detection kit directly measures the ROS produced by live cells; hence, low concentrations of AgNP and a short incubation time (6 hours) were used in the study represented in Figure 3b. Ag⁺ cytotoxicity can be caused by the induction of mitochondrial dysfunction through ROS production [51, 52]. The results shown in Figure 3c indicate that NAC, an ROS scavenger, could efficiently protect B16.F10 cells from the cytotoxic effects of Ag⁺ even when exposed to high concentrations of Ag⁺ (250 μg/ml). Pretreating melanoma cells with NAC prior to the addition of AgNP was also shown to be significantly protective, indicating the involvement of ROS production in promoting cell death in B16.F10 cells treated with AgNP. However, NAC failure to completely protect the cells indicated the possible involvement of ROS-independent cell death pathways such as autophagy [70].

In vivo antitumor effect of the combination of AgNP and RSQ on melanoma-challenged mice

RSQ as a monotherapy, when used as vaccine or treatment, has only modest effects on survival in animal melanoma models [19, 71]; however, when combined with tumor antigens [71] or a treatment that leads to the generation of tumor antigens the effects on survival can be synergistic [19]. In this study, AgNP were added as a tumoricidal agent with the aim of generating increased amounts of in situ melanoma antigens, that would then be available for uptake by DC. The combination of AgNP and RSQ was administered in one of two ways: either, RSQ was administered at the same time as AgNP (i.e. concurrent), or RSQ was administered 1 day after treatment with AgNP (sequential). The rationale behind including a sequential treatment was to allow time for AgNP to cause tumor cell death (thus creating a source of tumor antigens to be taken up by DC) prior to RSQ-induced induction of DC maturation (which leads to reduced phagocytic capacity [72]). The concentrations of AgNP and RSQ used in this in vivo study were based on concentrations used in previous studies where they were used independently and not in combination [73, 74]. Note that the AgNP concentration used in this experiment can be considered low compared to other similar work, where AgNP was subcutaneously injected into the mice [25].

The in vivo results showed that AgNP+RSQ (sequential) can significantly prolong the survival of melanoma-challenged mice compared to the control groups (naïve (untreated) and succinate-treated groups) (Figure 4). The tumor growth for individual mice and mean tumor growth per treatment group are shown in Figure 4b and Figure 4c, respectively. The tumor volumes at day 19 PTC were also compared and statistical analysis (Dunn’s multiple comparisons) revealed that the tumor volumes of mice treated with AgNP+RSQ (sequential) were significantly different from the naive (p-value of 0.0396, Figure 4d). From the survival curve (Figure 4e) and median survival data (Table S4), the mice in all the treatment groups at least trended toward surviving longer than the mice in the naïve and succinate-treated groups (vehicle). However, only mice treated with AgNP+RSQ (sequential) survived significantly longer than the naïve and succinate groups (p-values of 0.0103 and 0.0187, respectively) (Figure 4e and Table S5). The results presented here imply the importance of staggering the administrations of AgNP and RSQ administration times; however, there were no significant differences between treatment with AgNP+RSQ (sequential) and any of the other treatment groups. Further optimization may be required (e.g. increasing the interval between AgNP and RSQ treatments) to see if this sequential dosing strategy can be more efficient in terms of anti-tumor potency.

In vivo toxicity the combination of AgNP and RSQ on melanoma-challenged mice

Toxicity studies showed that the combination treatment tested in this study appeared to be non-toxic in mice as evidenced by: 1) no significant change in weights during the study period (Figure 5a); and 2) no significant increases in the levels of liver enzymes, alkaline phosphatase (ALP), aspartate aminotransferase (AST), and alanine aminotransferase (ALT) in the treatment groups (Figure 5b). A trend was seen for AST levels to be lower in all groups that involved RSQ treatment; however, it should be noted that the low levels of AST, still fell in the range found in normal mice (C57BL/6j) [75]. The low levels of ALT in the RSQ group were lower than the levels reported for healthy C57BL/6j mice [75]. Why this would be the case is not known. Hemolysis and lipemia indexes were also investigated. The results (table S6) showed normal levels for both indexes in all the treatments investigated, except for the hemolysis index in 1 mouse (1 out of 3 mice) in the RSQ treated group.

Figure 5: Toxicity study.

(a) Percent change in the mice weight. The mice weights were measured starting from day 7 post tumor challenge (PTC) and repeated every three days. These graphs showed the percent change in mice weight up to day 19 PTC as it was the last day all the naïve mice were still alive. Each data point shown here is the percent change in mouse weight between the day measured and three days before. Note that the tumor weight had been subtracted from the mice weight prior to the calculation (n = 9–10/treatment). (b) The level of alkaline phosphatase (ALP), aspartate aminotransferase (AST), and alanine aminotransferase (ALT) measured from the serum collected on day 21 PTC from the mice administered with different treatments (n = 3/treatment). The asterisk indicated significant differences in the level of the enzyme, compared to the naïve group. * indicated P-value ≤ 0.05.

The involvement of lymphocyte subsets in mediating the antitumor effect of AgNP+RSQ (sequential)

Since AgNP+RSQ (sequential) was the only treatment to significantly increase the survival of melanoma-challenged mice compared to the naïve group, we chose to deplete lymphocyte subsets to assess their potential importance in mediating the observed antitumor response. The combination of AgNP+RSQ (sequential) may mediate its antitumor effects in vivo through the activation of CD4⁺ or CD⁸⁺ lymphocytes [76, 77]. A survival study showed that there was trending difference between the treatment and the treatment +anti-CD8 (raw p-value of 0.0175 and adjusted p-value of 0.0817) (Figure 6, Table S7, Table S8). It should also be noted that the treatment was no longer significantly therapeutic (compared to the untreated/naive group) in the presence of anti-CD8 while it remained significantly therapeutic in the presence of anti-CD4 (Table S8).

NanoString analysis

NanoString technology was employed to screen for changes in expression of genes that may have been involved in the therapeutic mechanism of action of the combination of AgNP+RSQ (sequential). The Pan Cancer Immune Profiling panel was selected and was capable of detecting the expression of genes associated with immune responses (both innate and adaptive) and pathways of cell death. Here tumor tissue from the optimal treatment group (AgNP+RSQ (sequential)) was compared to tumor tissue from the untreated group.

Underlying mechanisms of the AgNP+RSQ (sequential) combination are still unknown, and instead of looking for the change in specific gene level between treated and untreated group, we looked for the overall changes in the relevance biological pathways. For this, we performed NanoString analysis and ran the data against Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and Gene Ontology (GO) term. The results (Figures 7 and 8) showed significant change in the gene expression in the immune response associated pathways (e.g. T cell activation, cell activation involved in immune response) and NF-κB pathway indicated the alteration in the immune and inflammatory states. The results also showed significant changes in the gene expression in the apoptosis-related and cell proliferation-related pathways (e.g. MAPK pathway). The list of the genes that showed significant differences expression levels between the two study groups are also shown in Table S9.

Figure 7: Nanostring results when ran against KEGG analysis pathways.

P.adjust = adjusted p-value. Color should be used in print.

Figure 8: Nanostring results when ran against the GO terms analysis pathways.

Noted that P.adjust = adjusted p-value. Color should be used in print.

The data presented here point toward candidate genes, whose expression may play a role in mediating the effects of the combination treatment. Validation via conventional methods such as qPCR, western blotting, and ELISA would still be needed to narrow down the salient contributors.

Conclusion

AgNP were synthesized using a chemical reduction method and their anti-melanoma effects were confirmed in vitro. The in vitro results indicated that AgNP can induce melanoma cell death through an apoptosis-related pathway. AgNP and RSQ (sequential) was shown to significantly increase the survival of melanoma-challenged mice, compared to the negative control groups and CD8+ T cells were implicated in being at least partially involved in the treatment mechanism. NanoString analysis revealed significant changes in gene expression in the cell proliferation, immune activation, and inflammatory pathways. The combination also appeared to be non-toxic in mice. The results from this study indicated that the combination of AgNP and RSQ may be promising for melanoma treatment and is worth further investigation.