Sublethal dosage for AuNPs exposure to macrophages
The properties of AuNPs with different diameters (5 nm or 50 nm) and coatings (BPEI, PVP, lipoic acid, tannic acid, citrate, and mPEG) were first determined. Fig. S1A shows that all AuNPs were approximately spherical and had uniform diameters. The hydrodynamic diameters of AuNPs detected by DLS were slightly higher than the diameters obtained from TEM (Fig. S1B), which may be caused by ligand adsorption and electrical double layers on the surface of the AuNPs. Most of the coated AuNPs were negative in both water and DMEM, wherein there was slightly more negative potential in water (Fig. S1C). The surface charge of BPEI-AuNPs was almost positive in the two types of media, and only 5 nm BPEI-AuNPs were negatively charged in DMEM. This phenomenon was regarded as the formation of protein coronas on particles, which leads to a change in zeta potential.
To determine the noncytotoxic concentration, the cell viability of macrophages treated with a series of concentrations of AuNPs was assessed. As shown in Fig. 1A-B and Fig. S2, none of the AuNPs showed an apparent alteration in cell viability when the concentration was at or lower than 5 µg/mL. At 20 µg/mL, except for mPEG-AuNPs, the cell morphologies were disturbed by AuNPs. At 5 µg/mL, there was no apparent morphological change upon treatment with AuNPs (Fig. 1C-D). Data from Annexin-V/PI dual staining further confirmed that AuNPs at 5 µg/mL did not cause dramatic cytotoxicity (Fig. S3). The noncytotoxic concentration selected in the study was also included in the range of expected environmental concentrations (1.6 to 16.6 µg/mL) in vitro, which was based on the calculation model of the National Institute for Occupational Safety and Health. Therefore, 5 µg/mL was considered the sublethal dose for RAW264.7 cells in the current study.
Many studies have indicated that the internal content of NPs is one of the pivotal factors influencing the interaction of NPs with organelles after exposure.[5, 13] ICP-MS was thus performed to evaluate the intracellular content of AuNPs. According to the calibration standard (Fig. S4), at the same external exposure dose of 5 µg/mL, the intracellular 5 nm and 50 nm BPEI-, PVP-, lipoic acid-, tannic acid-, citrate-, and mPEG-AuNPs were 0.463, 0.339, 0.366, 0.493, 0.451, 0.179, 0.501, 0.435, 0.509, 0.570, 0.603 and 0.110 pg/cell, respectively. The 5 nm and 50 nm mPEG-AuNP contents were lower than half of those of the other five AuNPs, with averages of 0.179 and 0.110 pg/cell, respectively. A series of studies have demonstrated that the cellular uptake of PEG-coated AuNPs was less than that of other coated AuNPs; for example, SK-BR-3 breast cancer cells preferred to take up AuNPs in the following order: poly (allylamine hydrochloride), anti-HER2 antibody and PEG. PEG is often used to reduce the uptake of NPs by macrophages.
However, except for mPEG-AuNPs, the other five types of AuNPs had little disparity (Fig. 1E). The similar intracellular content indicated that BPEI-, PVP-, lipoic acid-, tannic acid-, and citrate-coatings showed negligible effects on the cellular uptake of AuNPs. This fact might be attributed to the powerful phagocytic ability of macrophages toward NPs. Therefore, to compare the effect of different AuNP properties on macrophage mitochondria under sublethal doses and exclude the influence of internal content (Fig. 1E), BPEI-, PVP-, lipoic acid-, tannic acid-, and citrate-AuNP were used for further research.
Effects of AuNPs on the mitochondrial morphology and structure at sublethal dosages
To evaluate the influence of AuNPs with different diameters and coatings, the mitochondrial morphology and structure of RAW264.7 cells were evaluated after exposure for 24 h. As shown in Fig. S5, AuNPs altered the mitochondrial morphology to different degrees. Compared with the control, AuNP exposure reduced the mitochondrial fluorescence intensity and the number of tubular mitochondria, both of which showed an analogous downward trend (Fig. 2A-B). Of note, BPEI-AuNP always caused the maximum effects. The representative TEM images further demonstrated alterations in mitochondrial morphology and structure (Fig. 2C). For both 5 nm and 50 nm NPs, BPEI-AuNP, PVP-AuNP, lipoic acid-AuNP, and tannic acid-AuNP displayed higher swelling, vacuolization, and cristae fracture, as indicated by the green arrows. Taken together, these results showed that AuNPs with different diameters and coatings at sublethal doses can cause damage to mitochondrial morphology and structure. Among all of the AuNPs, BPEI-AuNPs caused a maximal effect on mitochondria.
Previous studies have indicated the different toxic effects of AuNPs on mitochondria. Karataş et al. found that AuNPs can form aggregates in the cytosol away from the mitochondria and did not cause substantial damage to mitochondria. However, in another study, AuNPs could be gradually trafficked to the mitochondria, where they reside in an aggregated state, making mitochondria somewhat swollen and round and causing mitochondrial crista to partially disappear and vacuolize. Our study serves as a proof-of-concept that both 5 nm- and 50 nm-coated AuNPs agglomerated in the lysosomes of the cytoplasm but not mitochondria. However, the intracellular AuNPs also caused swelling, vacuolization, and round-shaped mitochondria.
Impacts of AuNPs with different diameters and coatings on ROS generation and ATP content
The mitochondria have a central role in ATP production and ROS generation, and intracellular levels of ATP and ROS can reflect mitochondrial function to some extent.[31, 45] Therefore, the total ROS, mitochondrial ROS, and ATP content upon treatment with AuNPs were evaluated. As shown in Fig. S6, alterations in total ROS content were observed in AuNP-treated cells, and the degree of increased ROS was highly dependent on the diameter and coating. According to the quantitative results shown in Fig. 3A, the content of total ROS was significantly elevated (p < 0.01) after exposure to 5 nm BPEI-, PVP-, lipoic acid-, and tannic acid-AuNP. Furthermore, mitochondrial ROS was also increased (Fig. S7), especially in cells stimulated with 5 nm and 50 nm BPEI-, PVP-, lipoic acid-, tannic acid-AuNP, and 50 nm citrate-AuNP (Fig. 3B, p < 0.01). Similarly, many studies have shown that NPs could induce ROS production.[12, 46] For example, exposure of pristine graphene at sublethal doses (5 µg/mL) for 48 h induced ROS generation in RAW264.7 cells. Our results indicated that a low AuNP concentration could also cause an elevation of ROS in RAW264.7 cells. Moreover, NPs, including Au, Pt, TiO2, SiO2, and Fe2O3, increased the intracellular ROS of HUVECs at noncytotoxic concentrations, and the degree of change was closely related to the diameter and surface properties of the NPs. In addition, AuNPs could deplete the intracellular antioxidant pool, stimulate ROS production, and cause oxidative stress, finally leading to cell necrosis and apoptosis.[48, 49]
As shown in Fig. 3C, coated AuNP-treated cells showed significantly lower ATP content than untreated cells. ATP levels were reduced by 2.8, 2.1, 1.9, and 1.5 times after exposure to 5 nm BPEI-AuNPs, PVP-AuNPs, lipoic acid-AuNPs, and tannic acid-AuNPs for 24 h, respectively, compared with the control group (p < 0.01). The exposure of 50 nm BPEI-AuNPs, PVP-AuNPs, and lipoic acid-AuNPs also led to reduced ATP content (Fig. 3C, p < 0.01). Similar decreases in ATP after AuNP exposure have been presented in many previous studies, and the decrease was associated with cell cycle arrest.[50, 51] In the study of Yen et al., AuNPs ranging from 2 to 40 nm significantly inhibited the proliferation of J774A.1 macrophages, which further demonstrates that AuNPs may affect ATP production by macrophages through cell cycle arrest. Moreover, changes in anabolic states were also associated with the activation of proinflammatory macrophages, such as enhanced glycolytic metabolism and inhibited mitochondrial oxidative phosphorylation in inflammatory macrophages. The reduction in ATP caused by AuNPs suggests the inflammatory response of macrophages. Furthermore, the inhibition of ATP levels by AuNPs could be ascribed to the occurrence of apoptosis or necrosis, which is combined with changes in ROS levels. Although there was no significant change in cell activity under the noncytotoxic dose, the changes in ROS and ATP at the subcellular level still indicated the perturbed normal physiology status of the cell, further suggesting the potential harm of sublethal dose AuNPs with different diameters and coatings.
Correlation between mitochondrial alteration and AuNP properties
To determine the correlation of different properties and mitochondrial response and further elucidate the contribution of different properties to mitochondrial response at noncytotoxic dose AuNP exposure, cluster analysis, two-way ANOVA, and multiple linear regression were next performed. As shown in the heatmap of cluster analysis (Fig. 4A), the mitochondrial response was divided into two degrees. The alterations induced by 5 nm and 50 nm citrate-AuNP and 50 nm tannic acid-AuNP were classified into one degree, while those in the 5 nm and 50 nm BPEI-AuNP, PVP-AuNP, lipoic acid-AuNP, and 5 nm tannic acid-AuNP groups were classified into another degree. Surface properties led to the classification of two degrees, whereas the contribution of diameter was small. Further comparative analysis (shown in Additional file 2) illustrated that there were also differences between any two coatings with the same particle diameter for most of the mitochondrial indexes (p < 0.05). However, no obvious differences in most indicators were found between two diameters with the same coating. These results indicated that the surface properties exerted a higher impact on mitochondria in macrophages than the diameter, and final multiple linear regression (Table S1) further proved this conclusion.
According to the result of different coated AuNP characteristics (Fig. S1C), different coatings lead to an obvious change in zeta potential. Meanwhile, based on the result that the surface properties exerted a higher impact on mitochondria and the importance of zeta potential in surface properties, we further analyzed the correlation between zeta potential and mitochondrial response. The data from correlation analysis revealed that the mitochondrial response is related to the zeta potential (Table S2). With increasing zeta potential, the mitochondrial fluorescence intensity, the number of tubular mitochondria, and the ATP level decreased. In comparison, total ROS and mitochondrial ROS were increased (Fig. 4B). These data suggested that the differences in mitochondrial responses caused by different coatings may largely be related to the coating-related zeta potential alteration, and the higher zeta potential had a stronger effect on the mitochondria, indicating that the mitochondrial response induced by the sublethal dose of AuNP exposure might be surface charge-dependent.
Previous studies have concluded that the interactions of NPs with biological systems are responsible for the execution of NP functions and eventual toxicity.[5, 54] Though diameter and coating both could affect the interaction, diameter mainly influenced the cellular uptake pathways through a variety of diameter-dependent interactions with the lipid bilayer, while coating-related surface properties affected the membrane interactions through many kinds of approaches. Different from those unstable NPs, AuNPs is the most stable NPs in the ambient environment and even within cells for a long time, which is a proper model to investigate the effects of surface on its toxicity. Among the coating-related surface properties, the zeta potential of nanoparticles is one of the key factors. Many studies have shown that cells are more effective at uptaking positively charged AuNPs than negatively charged and neutral AuNPs.[5, 56] The reason for this phenomenon is that the cell membrane is mostly negatively charged so that AuNPs with greater zeta potential can be tightly combined and internalized to a greater extent than AuNPs with less zeta potential due to electrostatic interactions. Moreover, the membrane penetration ability of positively charged NPs was greater than that of neutral and negatively charged NPs, leading to a larger toxic response.[41, 58] However, our results showed that there was little difference in the intracellular contents of AuNPs with different coatings except for mPEG-AuNPs.
The response of mitochondrial respiratory chain complexes and macrophage function induced by typical AuNPs
Considering the obvious changes in ATP and ROS disturbed by AuNPs, two different coated AuNPs (i.e., BPEI-AuNPs and tannic acid-AuNPs) were selected to further study the alteration of the respiratory electron transport chain, a key site for ATP and ROS generation. As shown in Fig. 5A-B, there was no observable alteration in complex II (SDHB) or complex Ⅳ (COX Ⅱ) in macrophages stimulated by BPEI-AuNPs and tannic acid-AuNPs at both 5 nm and 50 nm. The expression of complex Ⅲ (UQCRC2) was downregulated in the 5 nm AuNP-exposed groups but not in the 50 nm AuNP groups when compared with the control group. However, there was a conspicuous decrease in complex Ⅴ (ATP5A) in macrophages treated with both 5 nm and 50 nm AuNPs. The activities of complex Ⅲ and complex Ⅴ were further determined. Interestingly, 5 nm and 50 nm BPEI- and tannic acid-AuNPs had no effect on the activity of complex Ⅲ in both RAW264.7 and J774A.1 macrophages (Fig. S8 A and C). However, a decreased activity of mitochondrial complex Ⅴ in the 5 nm AuNP-exposed groups was observed in both RAW264.7 and J774A.1 macrophages (Fig. S8 B and D), which was consistent with the results of western blots (Fig. 5A-B). The above data suggested that complex V of the mitochondrial respiratory chain was more sensitive to the exposure of AuNPs and can be listed as a meaningful biomarker for the high-throughput screening of NP-induced mitochondrial dysfunction at sublethal doses.
Some studies have indicated that NP exposure could cause changes in the mitochondrial respiratory chain.[59, 60] For example, NP exposure at a dose of 100 μg/mL downregulated the expression of mitochondrial respiratory chain complexes I, IV, and V in human bronchial epithelial (HBE) cells. However, research on the effects of NP exposure on the respiratory chain has mainly focused on high doses. Studies on the effects of NP exposure at sublethal doses on mitochondrial respiratory chains are lacking. One study indicated that at sublethal concentrations, the expression of complexes I, IV, and V was downregulated after NP exposure for 24 h. Therefore, the results of our study were a supplement to the data on the effects of NPs on mitochondrial respiratory chains at sublethal doses. In a series of mitochondrial respiratory chain complexes, complex III is one of the pivotal points in ROS generation, and complex V is the key enzyme that phosphorylates ADP to ATP. The downregulation of complex III and V might result in a decrease in ROS and ATP. In our study, the content of ATP was decreased after exposure to AuNPs with different coatings and diameters (Fig. 3C), which is consistent with the downregulation of complex V. Therefore, it was presumed that AuNPs downregulated complex V and further led to a decrease in ATP generation. Similarly, studies have indicated that NP exposure inhibited the expression of complex V and impaired ATP production, consistent with the results of our study.[59, 61] However, ROS generation was increased after exposure to different AuNPs (Fig. 3B), which contradicted the downregulation of complex III. The reason for this contradiction might be the other sources of ROS, such as complex I, NADPH oxidases, xanthine oxidase, and nitric oxide synthase. AuNPs may increase ROS production by influencing other sources of ROS. Alternatively, the increase in ROS might be related to the sensitivity of the methods used to detect mitochondrial ROS. Taken together, complex V might serve as a sensitive biomarker to indicate the effect on mitochondria under a low dose of NP exposure.
Subsequently, we examined the effects of AuNPs treatment on the biological functions of macrophages, such as the secretion of pro-inflammatory cytokines and phagocytic capacity of macrophages. The level of IL-6, a representative pro-inflammatory cytokine of macrophages, was significant increased in macrophages treated with 5 nm and 50 nm AuNPs when compared with the control group (Fig. S9). However, there was no difference in the ability of phagocytosis between 5 and 50 nm BEPI-, tannic acid-AuNPs treated groups and control group in both RAW264.7 macrophages and J774A.1 macrophages (Fig. S10). Similarly, a prior study by Chen et al. has reported that noncytotoxic dose of TiO2NPs treatment had no effects on phagocytic capability of RAW264.7 cells, but cytotoxic dose produced attenuation on that. Moreover, the TiO2NPs caused mitochondrial dysfunction and activated inflammatory responses under both the cytotoxic and noncytotoxic dose. Overall, these results indicated that AuNPs with different coating and size at a similar internal exposure dose under sublethal concentration led to the IL-6 mediated inflammation response, which might be linked to the changes in mitochondrial morphology, structure, and function.