Quantitative cytotoxicity analysis of antibacterial Janus nanoparticles in immune and cancer cells

Introduction

The rise of multi-drug resistant bacteria poses a significant threat to human health. In 2019 alone, antimicrobial resistance (AMR) caused an estimated 1.27 million deaths globally.¹ The AMR crisis is expected to worsen with the slowdown in new antibiotic development. To address the urgent need for effective antibacterial agents, nanoparticles (NPs) have emerged as a promising alternative to traditional antibiotics.^2–4 Owing to their unique physicochemical properties, antibacterial NPs are known to interact with bacterial cells through a combination of multiple mechanisms, including physical disruption of cell membranes and oxidative stress induction, which collectively minimize the likelihood of bacterial resistance development.^5–8

However, alongside their potential benefits, the cytotoxicity of antibacterial NPs, particularly their impact on mammalian cells, must be thoroughly evaluated, to ensure their therapeutic viability and safety. Previous research has shown that nanoparticle cytotoxicity can vary widely depending on their physiochemical properties, such as size, shape, surface charge, and composition.^9,10 For instance, NPs with certain surface chemistries, including positive charges and hydrophobicity, may induce oxidative stress or inflammation in mammalian cells, potentially leading to adverse effects.^11–14

Our previous work demonstrated that amphiphilic Janus NPs, designed with hydrophobic and polycationic ligands on separate hemispheres, are highly effective against both gram-positive and gram-negative bacteria.¹⁵ We showed that these Janus NPs effectively inhibit bacterial growth by physically disrupting the bacterial outer membranes and inducing oxidative stress in bacterial cells. Building upon these findings, the current study extends our investigation to assess the cytotoxicity of these NPs in mammalian cell cultures, focusing on immune cells and cancer cells. Using quantitative cytotoxicity assays with single-cell precision, we establish the relationship between nanoparticle concentration and cytotoxicity in vitro.

Methods

Results and Discussion

The amphiphilic Janus NPs (100 nm in diameter) have distinct surface coatings on two hemispheres (Figure 1A), same as we reported previously, and were fabricated using our established protocol.^15,16 Briefly, we first coated one hemisphere of aminated silica NPs with thin chromium (5 nm thickness) and gold layers (25 nm thickness) via directional vapor deposition, and then conjugated octadecanethiol on the gold cap to render it hydrophobic. The other hemisphere of the aminated silica NPs was conjugated with colistin using glutaraldehyde crosslinking chemistry. Colistin, typically reserved as a last-resort antibiotic, functions here as a polycationic ligand that enhances the strong electrostatic interaction of Janus NPs with bacterial membranes, as we reported previously.¹⁵

Figure 1. Schematic illustration of Janus nanoparticles (NPs) and the cell viability assay.

(A) Amphiphilic Janus NPs exhibit distinct surface makeups on two hemispheres. One side is conjugated with hydrophobic octadecane thiol ligands on a gold cap, while the other side is conjugated with the polycationic ligand colistin. (B) In the cytotoxicity assay, nuclei of all cells are stained with membrane permeable Hoechst dye (green). (C) Nuclei of non-viable cells with damaged plasma membrane are stained with the membrane impermeable dye propidium Iodide (red).

Based on our previous findings demonstrating the effective disruption of model lipid membranes and bacterial outer membranes by Janus NPs,^15,17–20 we proceeded to investigate their impact on plasma membrane integrity and subsequent cell viability. This investigation utilized a dual staining protocol involving propidium iodide and Hoechst, a widely accepted method for assessing cell cytotoxicity.^21,22 The assay was conducted in HeLa cells and RAW264.7 macrophage cells, two commonly used cell lines to represent cancer and immune cells, respectively. In this protocol, Hoechst dye, which penetrates intact membranes, stains the DNA of all cells, whether alive or dead (Figure 1B). Conversely, propidium iodide, a membrane-impermeable dye that binds to nucleic acids, selectively stains non-viable cells with compromised membranes. This approach allows us to not only identify cells with nanoparticle-induced plasma membrane permeabilization, but also quantify cytotoxic effects by comparing the number of propidium iodide-stained non-viable cells to the total Hoechst-stained cell population.

Initially, HeLa cells were incubated for 24h with Janus NPs at concentrations ranging up to 64 pM. The 0-64 pM concentration range was selected because it covers and exceeds the range of half maximal effective concentrations (EC50) of these NPs against four different bacterial strains (10-47 pM).¹⁵ We previously demonstrated that 64 pM Janus NPs induce substantial membrane damage and production of reactive oxygen species in almost all E. coli cells.¹⁵

Using both differential interference contrast (DIC) and widefield fluorescence microscopy, we observed that the morphology of HeLa cells remained unchanged in terms of both shape and size following exposure to Janus NPs (Figure 2A, C). There was a slight increase in the number of cells stained with propidium iodide, as nanoparticle concentration increased, indicating minor damage to the cell plasma membrane (Figure 2B). To quantify the impact on cell viability, we calculated the percent viability at each particle concentration using the equation: $\text{Percent Viability}=\frac{\mathit{\text{Number of Viable Cells}}}{\mathit{\text{Number of Total Cells}}}\times 100\%$. As shown in Figure 2D, even with 64 pM NPs, approximately 94.8% of cells remained viable. This reduction in cell viability is much less than the cytotoxic threshold defined by EN ISO 10993-5 as a reduction in cell viability by more than 30% compared to an untreated control.²³ Therefore, Janus NPs exhibited minimal effect on overall viability of HeLa cells.

Figure 2. Viability assay results of HeLa cells after interaction with varying concentrations of Janus NPs.

Differential interference contrast (DIC) (A) and widefield epi-fluorescence (B) images showing cells stained with Hoechst (green) and propidium iodide (red) after 24h incubation with Janus NPs at varying concentrations. (C) Zoomed-in DIC images of areas outlined in red rectangles in (A). Quantification of viability of HeLa cells as a function of Janus nanoparticle concentration (D). Each data point represents the average of two independent samples and each error bar represents the standard error between samples. Scale bars: 40 μm.

After assessing the impact of JNP treatment on HeLa cells, we investigated the cytotoxicity of the Janus NPs to professional phagocytes, a type of immune cells specialized in internalizing and digesting foreign particles including pathogens and synthetic NPs.^24–28 To this end, we evaluated the cytotoxicity of the NPs against RAW264.7 cells using the same dual staining protocol.

Using both DIC and widefield fluorescence imaging, we observed that some NPs were internalized by the macrophage cells, resulting in regions immediately adjacent to cells showing almost no NPs compared to background levels (Figure 3A, C). This phenomenon is more pronounced at higher nanoparticle concentrations. Meanwhile, the cell morphology changed slightly upon nanoparticle exposure. Post-interaction with NPs, cells exhibited increased heterogeneity in shape, characterized by enhanced membrane protrusions and asymmetry (Figure 3C).

Figure 3. Viability assay results of RAW264.7 macrophage cells after interaction with varying concentrations of Janus NPs.

Differential interference contrast (DIC) images (A) and Widefield epi-fluorescence images (B) showing cells stained with Hoechst (green) and propidium iodide (red) after 12h interaction with Janus NPs of varying concentrations. (C) Zoomed-in DIC images of areas outlined in red rectangles in (A). Quantification of viability of RAW264.7 macrophage cells as a function of Janus nanoparticle concentration (D). Each data point represents the average from three independent samples and each error bar represents the standard error between samples. Scale bars: 40 μm.

Consistent with these observations, there was a slight increase in the number of cells stained with propidium iodide at higher nanoparticle concentrations (Figure 3B), indicating concentration-dependent damage to the cell plasma membrane and subsequent cell death. As shown in Figure 3D, Janus NPs up to 16 pM had negligible effects on macrophage cell viability. However, cell viability decreased by 22.9% and 26.6% after exposure to 32 pM and 64 pM NPs, respectively. It is important to note that, even though higher concentrations of Janus NPs reduced cell viability, it did not fall below the threshold of a 30% reduction, which is considered cytotoxic as defined by the EN ISO 10993-5 standard.²³

By comparing the results from HeLa cells and RAW264.7 macrophage cells, it is evident that the cytotoxic effect of the Janus NPs is more pronounced in phagocytic immune cells, which inherently internalize a larger number of NPs. This is supported by viability quantification and our observation of changes in macrophage cell morphologies. This finding aligns with previous reports that internalization is a contributing factor in the cytotoxicity of NPs.^29–32

Importantly, neither cell type shows a significant cytotoxic effect from the Janus NPs, based on the established threshold of a 30% reduction in cell viability as indicative of toxicity. The nanoparticle concentrations we tested ranged from 0 to 64 pM, exceeding the EC50 values of the NPs against various bacteria (e.g., an EC50 of ~10 pM for E. coli).¹⁵ These results validate the potential of Janus NPs as antibacterial agents. This research opens doors to further studies, including more comprehensive assessments of both in vitro and in vivo cytotoxicity of the NPs using additional markers, such as the generation of reactive oxygen species and the inflammatory response of immune cells.

Author contribution statement

B.J., M.P., H.R., and S.B. carried out the experiments. All authors wrote the manuscript. Y.Y. conceived the project idea and supervised the project.

Ethics & consent

Ethical approval and consent were not required.