Bioinformatics methods for identifying differentially expressed genes and signaling pathways in nano-silica stimulated macrophages

Abstract

The incidence of disease relating to nanoparticle exposure has been rising rapidly in recent years, for which there is no effective treatment. Macrophage is suggested to play a crucial role in the development of pulmonary disease. To investigate the changes in macrophage after being stimulated by nanometer silica dust and to explore potential biomarkers and signaling pathways, the gene chip GSE13005 was downloaded from Gene Expression Omnibus database, which contained 21 samples: 3 samples per group and 7 groups in total. Macrophages in the control group were cultured in serum-free medium, while the experimental groups were treated with nanometer silica dust in different sizes and concentrations, respectively. To identify the differentially expressed genes and explore their potential functions, we adopted the gene ontology analysis and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis and also constructed protein–protein interaction network. As a result, 1972 differentially expressed genes were identified from 22,690 microarray data in the gene chip, 1069 genes were upregulated and 903 genes were downregulated. Results of the gene ontology analysis indicated that the differentially expressed genes were widely distributed in intracellular and extracellular regions, regulating macrophage apoptosis, inflammatory response, and cell differentiation. The Kyoto Encyclopedia of Genes and Genomes pathway analysis showed that the majority of differentially expressed genes were enriched in cytokine–cytokine receptor interaction, cancer or phagosome transcriptional misregulation. The top 10 hub genes, S100a9, Nos3, Psmd14, Psmd4, Lck, Atp6v1h, Jun, Foxh1, Pex14, and Fadd were identified from protein–protein interaction network. In addition, Nos3, Psmd14, Atp6v1h, and Jun were clustered into module M2 (r_c = 0.74, p < 0.01), which mainly regulates cell carcinogenesis and antivirus process. In conclusion, differentially expressed genes screened from this study may provide new insights into the exploration of mechanisms, biomarkers, and therapeutic targets for diseases relating to nanoparticle exposure.

Keywords

Lung cancer pneumoconiosis gene ontology differentially expressed genes macrophage

Introduction

With the wide application of nanomaterial, nanoparticle exposure has been becoming a common phenomenon in daily life. More and more studies have indicated that nanometer silica dust exposure is strongly associated with pulmonary inflammatory response, interstitial fibrosis, and lung cancer.^1–3 Similar to other dusts, nanoparticles involved in human body mainly depend on respiratory inhalation; due to the tiny size, the inhaled nanoparticles can hardly be exhaled. As a result, most of the particles accumulated in pulmonary alveoli, what is worse, a small amount of particles can directly go through the air–blood barrier, arriving in peripheral blood via pulmonary circulation and causing multi-organ disorders.^4,5 Current studies have found that alveolar macrophage (AM) plays a crucial role in foreign body phagocytosis. Consequently, silica dusts accumulated in alveoli would be identified and swallowed by AM, initiating the antigen-presenting process to immune cells.^6,7 Along with silica dust accumulation, it may induce AM apoptosis or autophagy, besides, the ruptured macrophages may release a variety of cytokines and chemokines into pulmonary interstitial space, activating pulmonary inflammatory response.^8,9 In addition, the stimulation induced by inflammatory factors further induces lung fibroblast transdifferentiating into myofibroblast. As a result, a large amount of collagen was synthesized and deposited in the extracellular matrix, which could subsequently promote pulmonary fibrosis.^10–12 Although studies at the cell level have preliminarily revealed the pathogenic process associated with nanoparticle exposure, the integrated mechanism still needs further exploration.

Advances in molecular biology and genomics make the further exploration possible, high throughput sequencing technology has been widely used in disease diagnosis and prognosis. Recently, some hub genes related to lung cancer have been approved, but nanoparticle-induced gene expression profile changes were rarely discussed. Waters et al.¹³ found that the gene expression profile of macrophages was highly conserved among stimulations induced by different particle sizes, but few differentially expressed genes (DEGs) were proposed limited by the old database and bioinformatics methods, and the protein–protein interaction (PPI) network was also deficient. However, the phagocytosis of nanoparticles by macrophage plays an important role in pneumoconiosis and lung cancer development, so it would be of great significance to identify DEGs in nano-silica exposed macrophages. In this study, we integrated the DNA microarray technology and bioinformatics technology comprehensively and subsequently identified the key DEGs and signaling pathways; furthermore, we explored the role of DEGs in pulmonary fibrosis development from three aspects, including cellular components, molecular function, and biological process, which were also used to mine potential pathogenic genes and biomarkers.

Materials and methods

Microarray data

Gene chip GSE13005 was downloaded from the Gene Expression Omnibus (GEO) database, which was based on Agilent GPL6480 platform (Affymetrix mouse genome 430A 2.0, Mouse430A_2), contained 21 samples: 3 samples per group and 7 groups in total. RAW 264.7 mouse macrophages in the control group were cultured in serum-free medium, while the experimental groups were exposed to 10 nm silica at 5 (low), 20 (mid), or 50 (high) µg/mL or 500 nm silica at 250 (low), 500 (mid), or 1000 (high) µg/mL for 2 h. As concluded by Waters et al.,¹³ the changes in the macrophages gene expression profile were highly conserved among different stimulations; here, we integrated all experimental groups as silica stimulating group and then made a comparison with the control group for the next step data mining.

DEGs screening and function exploration

First, SAS 9.2 was used to preprocess the raw data collected from GEO database, RMAExpress software was used to do background adjustment, quantile normalization, and summarization. After matching to the annotation file which was downloaded from the Affymetrix official website (http://www.affymetrix.com), the combined dataset was outputted, which included microarray data and official gene symbols. In the next step, we hierarchically clustered samples and genes respectively with the average linkage method, signal to noise was selected as the test-statistic assessment indicator, and three conditions were set for DEGs screening: false discovery rate (FDR) <20%, unfolded change >2, and p value <0.01. After logarithmic transformation, the DEGs expression correlationship between the control group and the experimental group was evaluated. Gene ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment were conducted under the preset conditions (thresholds account ≥2 and ease score ≤0.1) with Database for Annotation, Visualization and Integrated Discovery (DAVID) 6.8 online (https://david.ncifcrf.gov/).

PPI network and module clustering

Search Tool for the Retrieval of Interacting Genes (STRING) database is a biological database and web resource of known and predicted PPI which covers 184 million interaction pathways constructed by 9.6 million proteins from 2031 organisms. In this study, the combined score was set as 0.4 to screen DEGs, and the PPI dataset was clustered with the Molecular Complex Detection (MCODE) plug-in in Cytoscape 3.4.0; the cluster condition was set as follows: haircut mode, node score cutoff value 0.2, and K-core value 2. Moreover, all modules were screened with a cluster coefficient value ≥0.5 if not specified.

Results

1972 DEGs were identified

In total, 21 samples were included in the analysis: 3 samples in the control group and 18 samples in the experimental group. Data from each gene chip were merged into a consolidated database, the Gene-e software was used to identify DEGs, and three conditions were set for DEGs screening: FDR <20%, unfolded change >2 and p value <0.01. A total of 1972 genes were proposed after the analysis, of which, 1069 were upregulated and 903 were downregulated (Figure 1(c)). Moreover, the gene expression correlation analysis was conducted with the top 100 genes; the correlation coefficient was 0.9932 and shown in Figure 1(a) and (b) (p < 0.001).

Figure 1.

The screening results of differentially expressed genes (DEGs). (a) The correlation analysis between the experimental group and the control group; the top 100 DEGs were used to construct the correlation analysis model, the correlation coefficient was 0.9932 (p < 0.001). (b) The gene expression profile comparison between the experimental group and the control group, the value of mean, median, upper, and lower quartiles of the experimental group (red) were all higher than the control group (blue), respectively. (c) The heat map of the top 100 DEGs, the upregulated genes are shown in red, and the down regulated genes are shown in blue. Sample clustering result showed a good fit for the actual grouping method.

DEGs function exploration

DAVID provides a comprehensive set of functional annotation tools to understand biological meaning behind large list of genes. In this study, we used GO process to explore DEGs functions from three aspects, including cellular components, molecular function, and biological process. The results showed that DEGs were widely distributed in the whole cell and the extracellular regions, including exosomes and endocytic vesicles. Besides, the DEGs have a tight relationship with macrophage apoptosis, ion transport, inflammatory response, and cell differentiation, which also have an ability to bind proteins and function as cytokine or growth factor activators (Table 1). On the contrary, in order to understand the DEGs action mode, we adopted KEGG analysis, which was based on clustering similar genes into a same network, interestingly, the main action modes were cytokine–cytokine receptor interaction and transcriptional misregulation, the related signaling pathways were phosphoinositide 3-kinase (PI3K)-Akt, tumor necrosis factor (TNF), mitogen-activated protein kinase (MAPK), nuclear factor kappa B (NF-κB), RIG-I-like receptor, and Toll-like receptor (Table 2), which were consistent with previous studies in lung cancer.^14–16

Table 1.

Gene ontology analysis of differentially expressed genes related to nano-silica treated macrophages.

Category	Term/gene	Gene count	%	p value
Biological process	GO:0043066—regulation of cell apoptosis	86	4.6	3.1E-6
	GO:0006811—ion transport	100	5.3	5.0E-10
	GO:0006954—inflammatory response	69	3.7	1.2E-9
	GO:0045766—regulation of angiogenesis	34	1.8	7.1E-9
	GO:0030154—cell differentiation	109	5.8	5.7E-6
Cellular components	GO:0005576—extracellular region	280	15.0	1.1E-24
	GO:0016020—membrane	790	42.2	1.6E-20
	GO:0070062—extracellular exosome	355	17.9	5.8E-12
	GO:0031410—cytoplasmic vesicle	69	3.7	1.5E-2
	GO:0030139—endocytic vesicle	10	0.5	4.8E-2
Molecular function	GO:0005515—protein binding	540	28.9	6.2E-22
	GO:0005125—cytokine activity	52	2.8	2.0E-10
	GO:0005102—receptor binding	77	4.1	4.8E-9
	GO:0008083—growth factor activity	36	1.9	9.6E-8
	GO:0017134—fibroblast growth factor binding	12	0.6	8.7E-6

GO: gene ontology.

Table 2.

KEGG pathway enrichment analysis of differentially expressed genes related to nano-silica treated macrophages.

ID	Pathway	Gene count	%	p value	Genes
mmu04060	Cytokine–cytokine receptor interaction	53	2.83	9.44E-07	PDGFB, GDF5, IL13, TNFSF14, CXCR2, CXCR3, CXCL12, CXCL10, TNFRSF11B, CXCR6, IFNG, GHR, IL25, CCL21B, CCR5, PPBP, CD40LG, IFNB1, VEGFA, PDGFRB, IL12B, EDA, CSF3, CSF2, CCL3, CCL2, TNF, TNFRSF12A, CSF1, CRLF2, TNFRSF8, CD70, CX3CL1, CCL4, TNFRSF4, CCL7, LIF, TNFRSF1A, IFNA1, IL12RB1, IL23A, IFNA9, IL18R1, IL6, FLT1, MET, TNFSF9, CCL12, PRLR, CXCL13, CXCL16, IFNA11, IL5RA
mmu04151	PI3K-Akt signaling pathway	63	3.37	4.79E-05	FGF5, PDGFB, FGF15, FGF9, FGF14, EFNA2, LPAR1, NOS3, MYB, INSR, ANGPT2, GHR, SGK2, RELA, BCL2L11, DDIT4, G6PC, IFNB1, CHRM1, VEGFA, PDGFRB, MDM2, LAMC2, PPP2R2A, ITGA2B, FGFR2, CSF3, FGFR4, MCL1, CSF1, GNG13, ITGB4, ITGB3, GNG12, BCL2L1, LAMB3, IFNA1, ITGAV, ITGB6, TEK, GYS2, COL6A1, IFNA9, IL6, FLT1, MET, TCL1B4, NR4A1, COL5A3, TCL1B1, EPHA2, LAMA2, ITGA9, CDKN1A, ATF4, YWHAH, CD19, PRLR, TSC1, ITGA7, RHEB, IFNA11, INS2
mmu05202	Transcriptional misregulation in cancer	36	1.92	6.59E-05	CSF2, MMP9, CCNT1, HIST2H3C2, BCL2L1, ZBTB17, REL, ELK4, RUNX1, PLAT, TMPRSS2, NFKBIZ, IL6, FLT1, CEBPB, BAIAP3, RELA, MET, RXRG, SPINT1, AFF1, HMGA2, DDIT3, CDKN1A, CD86, EYA1, NUPR1, ID2, BCL2A1A, HIST1H3A, MDM2, HIST1H3D, PBX1, HIST1H3F, PLAU, KLF3
mmu04668	TNF signaling pathway	26	1.39	1.81E-04	CSF2, TNF, CCL2, PTGS2, CSF1, MMP9, EDN1, CXCL2, NFKBIA, CX3CL1, CXCL10, VCAM1, LIF, TNFRSF1A, MAP2K7, CFLAR, IL18R1, IL6, CEBPB, RELA, FADD, JUNB, CCL12, ATF4, JUN, SELE
mmu04010	MAPK signaling pathway	46	2.46	4.45E-04	FGFR2, FGF5, FGFR4, TNF, PDGFB, FGF15, FGF9, FGF14, DUSP10, CACNB1, MKNK1, CACNB3, GNG12, TNFRSF1A, BDNF, RAC3, ELK4, RASGRP4, MAP3K2, DUSP16, JUND, MAP2K7, HSPA8, NFATC1, LAMTOR3, RELA, RELB, CACNG5, NR4A1, CACNG3, FLNC, CACNA2D3, CACNA2D2, DDIT3, ATF4, DUSP2, DUSP1, JUN, GADD45G, PDGFRB, HSPB1, MAPK7, MAPK8IP1, GADD45B, DUSP8, CACNA1B
mmu04064	NF-κB signaling pathway	20	1.07	0.007	BCL10, CFLAR, TNF, PTGS2, RELA, RELB, TNFSF14, NFKBIA, CCL21B, BCL2L1, CXCL12, CCL4, VCAM1, TNFRSF1A, BCL2A1A, CD40LG, LCK, ZAP70, ERC1, PLAU
mmu04145	Phagosome	30	1.60	0.011	MBL2, MSR1, CD209G, ATP6V1B2, ITGB3, CD209B, ATP6V0E, STX12, STX18, ITGAV, TUBB6, ATP6V0D1, TUBA1A, H2-EA-PS, TUBB3, H2-M9, OLR1, H2-M10.1, ATP6V1H, ATP6V1D, H2-DMB2, ATP6V1F, ATP6V1C2, ITGB2L, H2-OA, FCGR2B, H2-BL, H2-EB1, RAB5A, ATP6V0A4
mmu04622	RIG-I-like receptor signaling pathway	14	0.75	0.028	OTUD5, TNF, ATG12, RELA, NFKBIB, NFKBIA, FADD, TANK, CXCL10, IFNA1, IFNB1, IFNA11, IFNA9, IL12B
mmu04620	Toll-like receptor signaling pathway	18	0.96	0.041	CCL3, IL6, TNF, RELA, NFKBIA, FADD, TLR5, CCL4, CXCL10, IFNA1, CD86, CD80, IFNB1, JUN, IFNA11, IFNA9, IL12B, MAP2K7

KEGG: Kyoto Encyclopedia of Genes and Genomes; PI3K: phosphoinositide 3-kinase; TNF: tumor necrosis factor; MAPK: mitogen-activated protein kinase; NF-κB: nuclear factor kappa B.

The PPI network construction

The PPI database was filtered using combined score >0.4, the top 10 hub genes, S100a9, Nos3, Psmd14, Psmd4, Lck, Atp6v1h, Jun, Foxh1, Pex14, and Fadd were identified from PPI network. Subsequently, records, which satisfied the filter criteria, were imported into Cytoscape for remodeling. Modules M1, M2, and M3, with a clustering coefficient ≥0.5, are shown in Table 3. The follow-up KEGG analysis pointed out that module M1 played an important role in neuroactive ligand–receptor interaction, cytokine–cytokine receptor interaction, calcium signaling pathway, and chemokine signaling pathway; module M2 mainly involved in the biosynthesis of proteasome and steroid, transcriptional misregulation, signaling pathways of TNF and hypoxia-inducible factor 1 (HIF-1); and module M3 had a close relationship with cell adhesion, arachidonic acid metabolism, PI3K-Akt, and Ras signaling pathways (Table 4 and Figure 2).

Table 3.

Module clustering of differentially expressed genes.

Module	Nodes (N)	Edges (N)	Average node degree	Score	r_c	p value
M1	57	939	32.9	33.536	0.93	<0.001
M2	133	1106	16.6	16.758	0.74	<0.001
M3	275	1517	11.0	11.073	0.50	<0.001

Table 4.

Pathways and differentially expressed genes in protein–protein interaction modules.

Module	ID	Pathway	Gene count	%	p value	Genes
M1	mmu04080	Neuroactive ligand–receptor interaction	24	42.11	4.49E-22	PTGER1, F2RL3, C3AR1, TACR2, GNRHR, BDKRB2, FPR2, LPAR1, PTGFR, NTSR2, GCGR, GRM1, SSTR2, HTR1A, HRH3, P2RY2, CNR1, CHRM1, TBXA2R, GRPR, ADRA1A, ADRA2C, HTR5A, HTR1F
	mmu04020	Calcium signaling pathway	11	19.30	4.62E-08	GNA14, PTGER1, TACR2, CHRM1, TBXA2R, GRPR, ADRA1A, BDKRB2, PTGFR, GRM1, HTR5A
	mmu04062	Chemokine signaling pathway	11	19.30	1.04E-07	PPBP, CCR5, CXCL13, CXCL16, CXCR6, CXCL2, CCL9, CXCR2, CXCR3, CXCL12, CXCL10
	mmu04060	Cytokine–cytokine receptor interaction	9	15.79	6.04E-05	PPBP, CCR5, CXCL13, CXCL16, CXCR6, CXCR2, CXCR3, CXCL12, CXCL10
M2	ssc03050	Proteasome	11	0.05	1.25E-10	PSMD14, PSMD13, PSMB6, PSMD12, PSMD11, PSMB3, IFNG, PSMC1, PSME3, PSMD4, PSMD8
	ssc04668	TNF signaling pathway	14	0.07	5.87E-10	CSF2, CFLAR, IL6, CEBPB, TNF, PTGS2, MMP9, RELA, NFKBIA, CX3CL1, JUNB, JUN, SELE
	ssc04066	HIF-1 signaling pathway	10	0.05	4.06E-06	CDKN1A, IL6, FLT1, RELA, TEK, IFNG, VEGFA, SERPINE1, NOS3, TIMP1
	ssc05202	Transcriptional misregulation in cancer	11	0.05	1.98E-05	CSF2, CDKN1A, IL6, FLT1, CEBPB, MMP9, RELA, RXRG, BCL2L1, PLAU
	ssc04721	Synaptic vesicle cycle	7	0.03	7.78E-05	ATP6V1C2, ATP6V1H, ATP6V1B2, ATP6V0A4, ATP6V0D1, ATP6V1D, ATP6V1F
	ssc00100	Steroid biosynthesis	5	0.02	1.17E-04	CYP51, SC5D, SQLE, FDFT1, NSDHL
M3	mmu04514	Cell adhesion molecules	19	7.01	1.81E-07	CLDN18, OCLN, CD8A, CLDN3, SELL, CLDN6, CLDN5, CLDN10, CDH2, PDCD1, VCAM1, CD86, CD80, CD40LG, H2-EB1, PECAM1, CLDN2, VCAN, CD4
	mmu00590	Arachidonic acid metabolism	13	4.80	2.86E-06	PTGES3, GGT5, PTGIS, GPX5, ALOX12E, CYP2B9, PLA2G12B, PLA2G1B, PLA2G6, GPX8, GPX7, ALOX8, PLA2G5
	mmu04151	PI3K-Akt signaling pathway	26	9.59	3.98E-06	FGF5, MCL1, PDGFB, ITGB4, ITGB3, LAMB3, IFNA1, IFNA9, ANGPT2, GHR, SGK2, MET, NR4A1, COL5A3, BCL2L11, LAMA2, ATF4, CD19, YWHAH, PRLR, ITGA7, IFNA11, MDM2, PDGFRB, LAMC2, ITGA2B
	mmu04060	Cytokine–cytokine receptor interaction	19	7.01	6.50E-05	CCL3, CCL2, PDGFB, MET, IL13, CCL4, LIF, TNFRSF1A, CCL12, IFNA1, IL23A, PRLR, CD40LG, PDGFRB, IFNA11, IFNA9, IL12B, IL5RA, GHR
	mmu04014	Ras signaling pathway	18	6.64	9.18E-05	FGF5, PDGFB, GRIN1, MET, FOXO4, PTPN11, REL, GAB2, RAC3, ETS2, GAB1, PLA2G12B, PLA2G1B, RAB5A, PDGFRB, PLA2G6, ANGPT2, PLA2G5

TNF: tumor necrosis factor; HIF: hypoxia-inducible factor; PI3K: phosphoinositide 3-kinase.

Figure 2.

The module M2 protein–protein interaction (PPI) network. The PPI database was filtered using combined score >0.4, then records, which satisfied the filter criteria, were imported into Cytoscape for remodeling. According to the preset screening criteria (haircut mode, node score cutoff value 0.2, and K-core value 2), three modules were returned, of which, M2 contains the largest number of hub genes (red).

Discussion

Long-term inhalation of productive dust may cause pulmonary fibrosis,¹⁷ which plays an important role in pneumoconiosis and lung cancer development. Since there is no effective treatment, the incidence of nanoparticle related disease has been increasing rapidly. As shown in many studies and reports, the number of new pneumoconiosis had reached 26,873 in 2014 in China,⁹ besides, over 3 million workers exposed to silica dust in India,¹⁸ about 2 million people have been working in dust exposure environment in America, of whom, the estimated death number of silicosis was 2500–5000 per year.¹⁹ Moreover, in the World Health Report of 2002, World Health Organization (WHO) estimated that 30,000 deaths and nearly 1,288,000 disability adjusted life years (DALYs) occurred yearly due to exposure to occupational airborne particulates.²⁰ Fortunately, current studies have revealed that the development of pneumoconiosis is particularly associated with alveolar inflammatory cell infiltration.²¹ Larson-Casey indicated that v-akt murine thymoma viral oncogene homolog 1 (Akt1) kinase–mediated mitophagy contributed to AM apoptosis resistance, which was required for pulmonary fibrosis.²² Other studies suggested that sustained PI3K activation exacerbated lung fibrosis via activating pro-inflammatory and pro-fibrotic pathways,²³ but gene level evidence was insufficient. Similarly, AKT1-PI3K signaling pathway was proposed as a critical part in the apoptosis of macrophage which was treated with nano-silica, it was found that 63 genes were differentially expressed, but most of them were seldom studied in this field. For example, Nos3, a crucial member of the AKT1-PI3K signaling pathway, was mainly discussed in cardiovascular or urinary system diseases but rarely in pulmonary fibrosis. GO analysis suggested that Nos3 is widely distributed in the nucleus, cellular membrane, Golgi, mitochondrion, and vesicles, which also participates in many biological metabolism processes, including macrophage apoptosis, endothelial cell migration, cell proliferation, signal transduction, nitric oxide biosynthesis, removal of superoxide radicals, and lung development. Therefore, it would be of great significance to explore the role of Nos3 in pulmonary inflammation and fibrosis, and Nos3 may also become a therapeutic target in pneumoconiosis in a long-term perspective.

As is known to all, nanoparticle is one of the most important risk factors for lung cancer, which was proposed in the 1980s, and subsequently confirmed as sufficient evidence in humans for the carcinogenicity by the International Agency for Research on Cancer (IARC) in 1996.²⁰ Turci held that the carcinogenesis of quartz dust did not depend on crystallinity but conchoidal crystal fragmentation, which could induce cellular membrane disruption and toxicity, resulting in macrophage inflammatory response.²⁴ In addition, the oxidized macrophage migration inhibitory factor (MIF) was found to be a promoter of cancer or tumor via AKT–extracellular signal–regulated protein kinases 1 and 2 (ERK1/2) signaling pathway.^25,26 Lim et al.²⁷ indicated that macrophages involved in pulmonary adenocarcinoma promoted tumor cell invasion and migration by upregulating the expression level of S100A8 and S100A9, which also accompanied by downregulating the matrix metalloproteinase 2 (MMP2) and MMP9. As a crucial component of DEGs, S100a8 and S100a9 are mainly distributed in the area of the plasma membrane, extracellular exosomes, and other regions, which have a capablility of binding metal ion, microtubules, and receptors. Moreover, S100a8 and S100a9 also take part in macrophage autophagy, apoptosis, leukocyte migration, inflammatory response, and innate immune response; however, both of them were not clustered in any models in this study, and the role of them still needs further mining and verification.

In addition to causing acute and chronic pulmonary diseases, nano-silica also induces autoimmune or autoimmune-related disorders, including scleroderma, systemic lupus erythematosus, rheumatoid arthritis, autoimmune hemolytic anemia, and dermatomyositis. Reports from the National Institute for Occupational Safety and Health (NIOSH) noted that there was an association between crystalline silica exposure with chronic renal diseases and sub-clinical renal changes.²⁰ Guo et al.²⁸ demonstrated an elevated messenger RNA (mRNA) level of collagen I and fibronectin in the heart and kidney after silica inhalation, but the severity of inflammatory response alleviated after anti-IL1β mAb treatment. In this study, we clustered the inflammatory response related signaling pathways in Table 2 and found that interleukin (IL13) might play an important role in macrophage response to silica stimulation. Interestingly, IL13 was also found to be an inducing factor in macrophage polarization in previous studies.^29–31

As an important member of Psmd families, Psmd14 participates in the synthesis and secretion of proteasome, but excessive proteasome induces extracellular matrix deposition.³² Atp6v1h, located in the lysosomal membrane and extracellular exosomes, was mainly involved in adenosine triphosphate (ATP) hydrolysis, synaptic vesicle cycle, and macrophage phagocytosis; however, available literatures for Atp6v1h only focused on its role in type 2 diabetes.^33,34 Jun exists in the nucleus as a proto-oncogene, regulating cell apoptosis and the activity of the transcriptional enzyme.³⁵ Although the molecular functions of the above hub genes are generally clear, their roles in nanoparticle exposure related diseases are seldom reported. In this study, most of the hub genes were integrated into module M2, and the M2 clustering coefficient was 0.74, showing a good fitting effect and suggesting M2 of great significance in disease mechanism exploration.

In summary, DEGs screened from this study may provide new insights into the exploration of mechanisms, biomarkers, and therapeutic targets for diseases related to nanoparticle exposure, but subsequent studies are still needed to validate the hypothesis.

Footnotes

Acknowledgements

L.Z. and C.H. contributed equally to this work.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Natural Science Foundation of China (81273039 and 81472954).

References

Vij

. Nano-based rescue of dysfunctional autophagy in chronic obstructive lung diseases. Expert Opin Drug Deliv 2017; 14: 483–489.

van den Brule

Beckers

Chaurand

. Nanometer-long Ge-imogolite nanotubes cause sustained lung inflammation and fibrosis in rats. Part Fibre Toxicol 2014; 11: 67.

Morimoto

Ogami

Todoroki

. Expression of inflammation-related cytokines following intratracheal instillation of nickel oxide nanoparticles. Nanotoxicology 2010; 4: 161–176.

Roberts

McKinney

Kan

. Pulmonary and cardiovascular responses of rats to inhalation of silver nanoparticles. J Toxicol Environ Health A 2013; 76: 651–668.

Bengalli

Mantecca

Camatini

. Effect of nanoparticles and environmental particles on a cocultures model of the air-blood barrier. Biomed Res Int 2013; 2013: 801214.

Hofmann

Blasche

Kasper

. A co-culture system with an organotypic lung slice and an immortal alveolar macrophage cell line to quantify silica-induced inflammation. PLoS ONE 2015; 10: e0117056.

Kim

Choi

. Cellular uptake, cytotoxicity, and innate immune response of silica-titania hollow nanoparticles based on size and surface functionality. ACS Nano 2010; 4: 5301–5313.

Liu

Pattie

Chandrasekara

. Epigenetic regulation of miRNA-124 and multiple downstream targets is associated with treatment response in myeloid malignancies. Oncol Lett 2016; 12: 2175–2180.

Wang

. Clinical statistics analysis on the characteristics of pneumoconiosis of Chinese miner population. J Thorac Dis 2016; 8: 2203–2211.

10.

Chen

Huang

. Mechanosensing by the alpha6-integrin confers an invasive fibroblast phenotype and mediates lung fibrosis. Nat Commun 2016; 7: 12564.

11.

Tzouvelekis

Lino Cardenas

. SH2 domain-containing phosphatase-SHP-2 is a novel anti-fibrotic regulator in pulmonary fibrosis. Am J Respir Crit Care Med 2016; 195: 500–514.

12.

Xie

Yang

. Sipa1 promoter polymorphism predicts risk and metastasis of lung cancer in Chinese. Mol Carcinog 2013; 52(Suppl. 1): E110–E117.

13.

Waters

Masiello

Zangar

. Macrophage responses to silica nanoparticles are highly conserved across particle sizes. Toxicol Sci 2009; 107: 553–569.

14.

Riquelme

Behrens

Lin

. Modulation of EZH2 expression by MEK-ERK or PI3K-AKT signaling in lung cancer is dictated by different KRAS oncogene mutations. Cancer Res 2016; 76: 675–685.

15.

Guo

Zhu

Zhao

. Tumor-associated calcium signal transducer 2 regulates neovascularization of non-small-cell lung cancer via activating ERK1/2 signaling pathway. Tumour Biol 2017; 39: 1010428317694324.

16.

Yue

Zhou

Gao

. High mobility group box 1/toll-like receptor 4/myeloid differentiation factor 88 signaling promotes progression of gastric cancer. Tumour Biol 2017; 39: 1010428317694312.

17.

Zhang

. N-acetylcysteine alleviated silica-induced lung fibrosis in rats by down-regulation of ROS and mitochondrial apoptosis signaling. Toxicol Mech Methods 2014; 24: 212–219.

18.

Jindal

. Silicosis in India: past and present. Curr Opin Pulm Med 2013; 19: 163–168.

19.

Thomas

Kelley

. A brief review of silicosis in the United States. Environ Health Insights 2010; 4: 21–26.

20.

Fedotov

. The ILO/WHO global programme for the elimination of silicosis (GPES). Afr Newsl Occup Health Saf 2005; 15: 56.

21.

Imazu

Yanagi

Miyoshi

. Ghrelin ameliorates bleomycin-induced acute lung injury by protecting alveolar epithelial cells and suppressing lung inflammation. Eur J Pharmacol 2011; 672: 153–158.

22.

Larson-Casey

Deshane

Ryan

. Macrophage Akt1 kinase-mediated mitophagy modulates apoptosis resistance and pulmonary fibrosis. Immunity 2016; 44: 582–596.

23.

Kral

Kuttke

Schrottmaier

. Erratum: sustained PI3K Activation exacerbates BLM-induced Lung Fibrosis via activation of pro-inflammatory and pro-fibrotic pathways. Sci Rep 2016; 7: 26048.

24.

Turci

Pavan

Leinardi

. Revisiting the paradigm of silica pathogenicity with synthetic quartz crystals: the role of crystallinity and surface disorder. Part Fibre Toxicol 2016; 13: 32.

25.

Olivieri

Bargagli

Inghilleri

. Macrophage migration inhibitory factor in lung tissue of idiopathic pulmonary fibrosis patients. Exp Lung Res 2016; 42: 263–266.

26.

Schinagl

Thiele

Douillard

. Oxidized macrophage migration inhibitory factor is a potential new tissue marker and drug target in cancer. Oncotarget 2016; 7: 73486–73496.

27.

Lim

Yuzhalin

Gordon-Weeks

. Tumor-infiltrating monocytes/macrophages promote tumor invasion and migration by upregulating S100A8 and S100A9 expression in cancer cells. Oncogene 2016; 35: 5735–5745.

28.

Guo

Shi

Cui

. Effects of silica exposure on the cardiac and renal inflammatory and fibrotic response and the antagonistic role of interleukin-1 beta in C57BL/6 mice. Arch Toxicol 2016; 90: 247–258.

29.

Yang

Ning

. Cigarette smoke extract-treated mast cells promote alveolar macrophage infiltration and polarization in experimental chronic obstructive pulmonary disease. Inhal Toxicol 2015; 27: 822–831.

30.

Smith

Tse

Read

. Regulation of macrophage polarization and plasticity by complex activation signals. Integr Biol 2016; 8: 946–955.

31.

Zhou

Huang

Lin

. Macrophage polarization and function with emphasis on the evolving roles of coordinated regulation of cellular signaling pathways. Cell Signal 2014; 26: 192–197.

32.

Yerlikaya

Okur

Baykal

. A proteomic analysis of p53-independent induction of apoptosis by bortezomib in 4T1 breast cancer cell line. J Proteomics 2015; 113: 315–325.

33.

Yang

Allister

. The identification of potential factors associated with the development of type 2 diabetes: a quantitative proteomics approach. Mol Cell Proteomics 2008; 7: 1434–1451.

34.

Molina

Rentfro

. Decreased expression of ATP6V1H in type 2 diabetes: a pilot report on the diabetes risk study in Mexican Americans. Biochem Biophys Res Commun 2011; 412: 728–731.

35.

Riesenberg

Groetchen

Siddaway

. MITF and c-Jun antagonism interconnects melanoma dedifferentiation with pro-inflammatory cytokine responsiveness and myeloid cell recruitment. Nat Commun 2015; 6: 8755.