Global gene expression analysis of macrophage response induced by nonporous and porous silica nanoparticles

Abstract

Little is known about the global gene expression profile of macrophages in response to changes in size and porosity of silica nanoparticles (SNPs). Spherical nonporous SNPs of two different diameters, and mesoporous spherical SNPs with comparable size were characterized. Reactive oxygen species, mitochondrial membrane potential, lysosome degradation capacity, and lysosome pH were measured to evaluate the influence of nonporous and mesoporous SNPs on mitochondrial and lysosomal function. RNA-sequencing was utilized to generate transcriptional profiles of RAW264.7 macrophages exposed to non-toxic SNP doses. DESeq2, limma, and BinReg2 software were used to analyze the data based on both unsupervised and supervised strategies to identify genes with greatest differences among NP treatments. Utilizing GATHER and DAVID software, possible induced pathways were studied. We found that mesoporous silica nanoparticles are capable of altering gene expression in macrophages at doses that do not elicit acute cytotoxicity, while gene transcription was minimally affected by nonporous SNPs.

Keywords: Gene expression; Gene ontology; Lysosome pathway; Nanotoxicity; Reactive oxygen species; Silica nanoparticles.

Figure 1

Scanning electron microscopy images of (A) Stöber SNPs with average diameter of 46±4.9 nm(Stöber50), (B) Stöber SNPs with average diameter of 432±18.7 nm (Stöber500), (C) mesoporous SNPs with average diameter of 466±86 nm (MSN500), and (D) PEGylated mesoporous SNPs with average diameter of 466±86 nm (MSN500-PEG). Transmission electron microscopy images of (E) Stöber50, (F) Stöber500, (G) MSN500, and (H) MSN500-PEG.

Figure 2

Raw264.7 macrophages exposed to sub-cytotoxic doses of SNPs with diverse physicochemical properties showed different gene expression responses. (A) Heat map of the gene expression response generated based on RNA sequencing revealed no genomic alteration for nonporous SNPs, independent from their size, however, mesoporous SNPs and PEGylated mesoporous SNPs elicited 192 and 136 gene expression changes compared to control, respectively. The heat map provides the Euclidean distances in gene-space with red rectangles representing minimal expression differences in gene-space, while blue rectangles indicate more extensive expression differences in gene-space. (B) BinReg2 analysis evaluating the probability of segregating the geneset. The scatter plot shows the predictions from the signature for each sample. On the Y-axis, high probabilities indicate that the gene expression profile of the sample better resembles the mesoporous group, while low probabilities indicate a closer resemblance to control and Stöber group. The blue and red circles are the predictions (from leave-one-out cross-validation) on the control and mesoporous samples, respectively. The error bars show the 95% credible interval. The X-axis, the Metagene Score, is the magnitude of the sample on the first principal component. This is used to help separate the samples across the plot. (C) The top 20 differentially expressed genes between SNPs and control groups in RAW264.7 macrophages. Darker red indicates increased gene expression (fold change) for each gene, white colors indicates no expression alteration, and darker blue color indicates lower gene expression (fold change) for each gene corresponding to each specimen.

Figure 3

Raw264.7 macrophages were exposed to half LC₅₀ dose of MSN500 for 4h, and RNA sequencing was used to determine gene expression profile compared with control samples. Genes with at least 2 fold changes were deposited in the DAVID and GATHER on-line database to generate gene ontology classification based on up-regulation or down-regulation. Statistically significant categories were determined considering genes with presence of at least 2 genes in each category, p<0.01, and Bayes factor over 4. (A) Functional categories for up-regulated genes. (B) Functional categories for down-regulated genes. Correlating genes of each category is shown in parentheses. (C) KEGG pathway analysis demonstrated central part of lysosome pathways up-regulation by the MSN500 SNPs in RAW264.7 cells. Genes with expression changes are highlighted with the asterisks.

Figure 4

qPCR analysis confirmed that by treatment of macrophages with mesoporous MSN500 SNPs, gene expressions of Atp6v0d2, Slc39a2, Tnfsf13, and Hvcn1 were significantly increased. Data are expressed as mean ± S.D. from (n = 3), ** P <0.01 and *** P <0.001.

Figure 5

(A) A Click-iT EdU assay was utilized to incorporate EdU to DNA during synthesis by Click chemistry. Fluorescent detection correlated with rate of cell proliferation. While all concentrations could be considered sub-cytotoxic for MSN500, nonporous Stöber500 at 112 μg/ml concentrations induced acute cytotoxicity in RAW264.7 macrophages. (B) A luminescence-based ATP concentration determination assay showed increased ATP synthesis for treated cells compared against non-treated control samples. Data are expressed as mean ± S.D. from (n = 3), ** P <0.01 and *** P <0.001.

Figure 6

ROS production and disruption of MMP induced by SNPs. (A) FCM quantification of the fluorescence emitted from DCFH-DA probe upon oxidation for the detection of intracellular H₂O₂. Treatment of macrophages did not induce any significant intracellular ROS. (B) Mitochondrial ROS measurement using MitoSOX™ via FCM. Statistically significant Mitochondrial ROS production were detected only for nonporous Stöber500 at 112μg/ml. (C) The MMP quantification using the JC-1fluorescent probe. Nonporous Stöber500 at 112μg/ml concentration results in disruption of MMP, but other concentrations for both nonporous and mesoporous particles do not disrupt the MMP. JC-1 fluoresces red when the mitochondrion is normally polarized, while green means mitochondrial potential has been depolarized. Positive control carbonyl cyanide 3-chlorophenylhydrazone (CCCP) at 25μM disrupts the MMP. CCCP, and Nonporous Stöber500 at 112μg/ml showed significantly different MMP loss compared to the control samples. (D) The representative FCM scatter plots of JC-1. Green fluorescence of depolarized mitochondrial potential is distributed in the bottom right-hand area, while red fluorescence of normally polarized mitochondrion is distributed in the upper right-hand area. Data are expressed as mean ± S.D. from (n = 3), ** P <0.01 and *** P <0.001.

Figure 7

Lysosome degradation capacity and lysosome pH induction by porous and nonporous SNPs. Stober17, Stober33, Stober52, Stober112, MSN17, MSN33, MSN52, and MSN112 correlates to Stober500 and MSN500 SNPs treatment with increased concentration from 17 to 112 μg/ml. (A) DQ-BSA analysis of lysosomal activity by confocal microscope imaging. RAW264.7 macrophages were incubated for 4 h with medium as control or with different concentrations of SNPs. Control sample presented bright fluorescent fragments, demonstrating active lysosomal degradation capacity of DQ-BSA. Fluorescent signal diminished by increased concentrations in both types of SNPs indicating decreased lysosomal proteolysis activity. (B) Fluorescence intensity quantified by densitometry (IOD) from at least 45 cells in each replicate (n=3). (C) Impact of SNPs treatment on lysosome pH measured by flow cytometry using LysoSensor Green DND-189 revealed dose dependent lysosome alkalinization only by mesoporous particles. (D) Lysosensor fluorescence intensity overlap in which Control is grey, MSN112 is the dashed line, and Stober112 is the solid line. Shift to lower intensities, indicates decrease of lysosome pH. Statistically significant differences were observed in toxic concentrations of Stöber500 and sub-cytotoxic concentrations of MSN500. Data are expressed as mean ± S.D. from (n = 3). *** P <0.001.

Figure 8

TEM images of RAW264.7 cells treated with media (control) or treated with different concentrations of porous and nonporous SNPs for 4 h. The SNPs were taken up by cells and localized inside vesicles. Dose-dependent increase of cellular association of SNPs in both types of nanoparticles is visualized. Red arrows point particles inside cells. Particles were not observed inside the nucleus.