Genome-Wide mRNA Profiling Identifies  the NRF2- Regulated Lymphocyte Oxidative Status in Patients with Silicosis


 BackgroundThe immunomodulatory abnormalities of silicosis are related to the lymphocyte oxidative state. The potential effect of antioxidant therapy on silicosis may depend on the variation in nuclear factor erythroid 2-related factor 2 (NRF2)-regulated antioxidant genes in peripheral blood mononuclear cells (PBMCs). As NRF2 is a redox-sensitive transcription factor, its possible roles and underlying mechanism in the treatment of silicosis need to be clarified.MethodsIn this study, the lymphocyte oxidative state was investigated by evaluating NRF2 expression and NRF2-dependent antioxidative genes in PBMCs from patients with silicosis. Key differentially expressed genes (DEGs) and signaling pathways were identified utilizing RNA sequencing (RNA-Seq) and bioinformatics technology. Gene set enrichment analysis was used to identify the differences in NRF2 signaling networks between patients with silicosis and healthy controls.ResultsThe number of monocytes increased significantly through cytology detection in patients with silicosis. Furthermore, RNA-Seq findings were confirmed using quantitative polymerase chain reaction and revealed that NRF2-regulated DEGs were associated with glutathione metabolism, transforming growth factor-β, and the extracellular matrix receptor interaction signaling pathway in PBMCs from patients with silicosis. The top 10 hub genes were identified by PPI analysis: SMAD2, MAPK3, THBS1, SMAD3, ITGB3, integrin alpha-V (ITGAV), von Willebrand factor (VWF), BMP4, CD44, and SMAD7.ConclusionsThese findings suggest that NRF2 signaling regulates the lymphocyte oxidative state and may contribute to fibrogenic responses in human PBMCs. Therefore, NRF2 might serve as a novel preventive and therapeutic candidate for silicosis.


Introduction
Silicosis is a pulmonary interstitial brosis disease caused by the exposure to crystalline silica dust. This disease is a progressive process characterized by lung in ammation during the early phase and is almost always fatal [1]. Unfortunately, no effective treatments are available, except the possibility of a lung transplant in a small minority of patients [2]. At present, a large number of employees are exposed to high concentrations of dust in a booming mining industry in China. The number of new silicosis and pneumoconiosis in coal worker cases reported per year reached more than 20,000 in China [3].
Silicosis is characterized not only by its direct brotic effect on lung tissue but by immunomodulatory abnormalities, such as the appearance of complications of autoimmune diseases and autoantibodies in silica-exposed populations [4][5][6]. Peripheral blood mononuclear cells (PBMCs) are collected from the peripheral or circulating blood and possess a single, round nucleus. PBMCs are the major immune cells in the human body and provide selective responses to the immune system. The roles of these blood cells are to adapt to intruders and ght infection. As a mixed population of single nuclear cells, PBMCs mainly comprise T-lymphocytes, B-lymphocytes, and natural killer (NK) cells and are also a rich source of monocytes, such as macrophages and dendritic cells [7]. PBMCs play a crucial role in maintaining and controlling the immunomodulatory process in silicosis pathogenesis [8]. The alteration of gene expression in PBMCs derived from patients with silicosis causes dysregulation of apoptosis and the acquisition of abnormal autoimmunity [9,10]. Animal experiments demonstrated that silica exposure induces migration of dendritic cells from the peripheral blood into the alveoli in rats [11]. Taken together, PBMCs are associated with the pathogenesis or development of silicosis.
Oxidative stress occurs due to a persistent imbalance of redox homeostasis induced by overproduction of reactive oxygen species (ROS) in silica-exposed populations. When ROS production exceeds antioxidant capacity, oxidative stress may have harmful effects on the structure and function of biological tissues. It is essential for oxidative damage to occur in the development of silicosis [12][13][14]. It was showed that the antioxidant N-acetylcysteine alleviates lung brosis induced by silica in rats by downregulating ROS and interfering with apoptosis signaling in our previous study [15]. Antioxidant treatments have been shown to be effective for silicosis in animal experiments [16][17][18]. However, antioxidant therapy has produced con icting results in clinical studies of the treatment of pulmonary brosis and other chronic diseases [19,20]. The failure of clinical trials may be due to the lack of understanding of the role of ROS in the development of silicosis. Indeed, clinical effects of antioxidant therapy on silicosis depend on the redox microenvironment, which changes spatially and temporally in different cell types and in different subcellular compartments [21]. Thus, the individual's difference in redox microenvironment potentially depends on variation of their antioxidant genes. We hypothesize that the endogenous antioxidant levels in the subjects might have greatly affected the results of the clinical trials.
Antioxidant defense systems to combat ROS from tissues oxidative damage include enzymatic antioxidants and nonenzymatic antioxidants. Enzymatic antioxidants comprise superoxide dismutases (SODs), catalase (CAT), glutathione peroxidases (GPXs), glutathione reductase (GR), and aldehyde dehydrogenases (ALDH). Glutathione (GSH), coenzyme Q10, ascorbic acid (vitamin C) and α-tocopherol (vitamin E) represent nonenzymatic antioxidants, which in uence each other and have their own unique roles [22]. The GPx/GR antioxidant system is related to other antioxidant systems, including the SOD/CAT system. GSH modulates the neutralization of free radicals by vitamin C (ascorbic acid) and vitamin E [23]. Occupational silica exposure leads to the alteration of SOD and CAT activities in the antioxidant defense system [24].
Nuclear factor erythroid 2-related factor 2 (NRF2) is a redox-sensitive transcription factor and plays a critical role in maintaining redox homeostasis by regulating the expression of antioxidant defense enzymes [25]. Levels of NRF2 in PBMCs increase immediately after ozone/oxygen exposure [26]. In a human intervention study, antioxidant modulated the NRF2-related gene expression in PBMCs [27]. Our previous research shows that NRF2 is involved in mediating the development of silicosis in animal experiments [28]. Although studies in animals have preliminarily revealed that the pathogenic process of silicosis associated with oxidative stress is regulated by NRF2, the integrated mechanism requires further exploration. Advances in molecular biology and genomics have enabled further exploration as highthroughput sequencing technology has been widely used in disease diagnosis and prognosis. However, gene expression pro le changes have rarely been discussed in PBMCs of patients with silicosis.
In the present study, genome-wide mRNA pro ling was used to identify the lymphocyte oxidative state regulated by NRF2 in patients with silicosis, and to investigate whether there was an association between lymphocyte oxidative state and the pathogenesis of silicosis. RNA sequencing (RNA-Seq) technology and bioinformatics technology were comprehensively integrated and subsequently identi ed key differentially expressed genes (DEGs) and signaling pathways. The role of DEGs in silicosis development was explored from three aspects, including cellular components, molecular function, and biological process, which were also used to mine potential pathogenic genes and biomarkers. Furthermore, the clinical signi cance of identi ed genes was veri ed using clinical samples. These data provide novel information and further understanding of the mechanism underlying oxidative stress in PBMCs from patients with silicosis.

Study design and participants
A total of 92 male patients with silicosis (average age: 40-65 years) and 87 male healthy volunteers (average age: 40-65 years) were included in the study. The healthy controls group had no dust exposure in the past. The group of patients with silicosis were strictly selected according to the standard diagnoses of pneumoconiosis (GBZ70-2015), and the patients were diagnosed as silicosis stage I. All subjects had no clinical symptoms of autoimmune disease, including Raynaud's phenomenon, sclerotic skin, arthralgia or facial erythema, or malignant tumors. Additionally, all subjects were negative for pneumonia, active pulmonary tuberculosis, pulmonary heart disease, infectious diseases and other lung related diseases.
All subjects con rmed that they understood the experimental procedure and provided written informed consent. The study was approved by the Ethics Committee of the Xinxiang Medical University (protocol number XYLL-2017086,approved 3 March 2017).
In this study, information was collected using selected survey questionnaires by trained medical personnel. The survey mainly included information related to age, smoking status, basic health status, and family history. Physical examinations included blood pressure, weight, and height measurements to determine body mass and body mass index (BMI). Levels of PBMCs in every subject were detected by an automated hematology analyzer (Cell-Dyn Sapphire, Abbott, USA). Pulmonary function evaluation was performed using a portable handheld spirometer (Drägerwerk AG, Lűbeck, Germany). The general speci cations for the performance, as well as interpretation, of the pulmonary function test were followed [29,30]. The pulmonary function indices documented were the forced vital capacity (FVC) and forced expiratory volume in one second (FEV1). All readings were in units of liters and percentage predicted values. The transforming growth factor (TGF)-β1 content was detected in accordance with manufacturer instructions by ELISA. In addition, three patients with silicosis and three healthy donors were randomly selected for the detection of genome-wide mRNA pro ling. 36 participants were respectively selected from the remaining healthy and patient samples with random and double-blindness method for mRNA analysis, immuno uorescence and western blot analysis.

PBMC isolation
PBMCs were isolated from 10 ml of fresh blood from patients with silicosis and healthy controls within 30 min after collection using the Ficoll-Paque (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA) density centrifugation method. Brie y, blood was diluted 1:1 with RPMI-1640 medium at room temperature (Solarbio, Beijing, China), underlaid with 10 ml Ficoll-Paque, and then centrifuged (1000 × g, 10 min, 20°C) with acceleration set to six and deceleration set to zero in a Heraeus Multifuge X3R (Thermo Fisher Scienti c, Langenselbold, Germany). Separated PBMCs were carefully collected and transferred to a 50 ml conical tube, which was brought up to 50 ml with RPMI-1640 medium, and then washed three times with RPMI-1640 medium (100 × g, 10 min, 4°C) with acceleration set to nine and deceleration set to six. The supernatant was removed and the pellet was resuspended in 1.0 ml RPMI-1640 medium for total RNA extraction, immuno uorescence assays or western blotting. PBMCs were enumerated and assessed for viability using trypan blue. Samples were stored at -80°C before use.
Total RNA extraction, library construction, and sequencing.
Total cellular RNA was extracted from PBMCs using the TRIzol reagent (Gibco BRL Life Technologies, Gaithersburg, MD, USA), according to the manufacturer's instructions. Extracted total RNA was analyzed using the Experion™ RNA StdSens kit (BIORAD, Gladesville, NSW, Australia) to con rm its quality and quantity. An mRNA library was constructed using the NEB Next Ultra RNA library prep kit (New England Biolabs, Ipswich, MA, USA). An Agilent 2100 (Agilent, Santa Clara, CA, USA) was used for library quality inspection and an Illumina HiSeq TM 4000 (Illumina, Santa Clara, CA, USA) was used for sequencing. mRNA sequencing was performed by Gene Denovo Biotechnology Co. (Guangzhou, China). To ensure data quality, the original data was ltered to reduce data noise before analysis.
Genome-wide mRNA pro ling DEG analysis Bioinformatics was used to reanalyzed the original data from the sequencing results. Principal component analysis was performed using the R statistical software package with normalized counts to investigate if samples from the same group clustered together. The R-based software package edgeR was used to determine whether there was a signi cant difference in gene expression levels between the two groups. Genes with a fold change ≥ 2 and P < 0.05 in a comparison were classi ed as signi cant DEGs. Functional analysis of DEGs was performed using the gene ontology (GO) project (http://www.geneontology. org) [31]. Pathway enrichment analysis was performed to identify signi cant pathways involving DEGs using the KEGG database [32].

Co-expression network analysis
To reveal functional associations between proteins, the STRING database was used to construct a PPI network [34]. In the PPI network, each node represents a protein and each edge (between two nodes) represents an interaction between these two proteins. Hub proteins were the nodes with a relatively large number of edges. Within the network analysis, the degree of association is an important factor to determine the relative importance of a gene. Different colors and sizes of nodes were employed to discriminate the degree of associations for one gene with the surrounding nodes. The co-expression networks were constructed using Cytoscape [35].

Quantitative real-time polymerase chain reaction (RT-qPCR) validation of responses in patients with silicosis
RT-qPCR analysis was performed as previously described [18]. Total RNA was extracted from PBMCs of healthy controls and patients with silicosis using the TRIzol reagent (Invitrogen, San Diego, CA, USA) and reverse-transcribed by the Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Scienti c, USA).
The relative expression levels of mRNAs encoding NRF2, Kelch-like ECH-associated protein 1 (KEAP1), glutamate-cysteine ligase modi er subunit (GCLM), TGF-β, mothers against decapentaplegic homolog 3 (SMAD3), SMAD2, mitogen-activated protein kinase 3 (MAPK3), thrombospondin 1 (THBS1), integrin beta-3 (ITBG3), bone morphogenetic protein 4 (BMP4), CD44 molecule (Indian blood group) (CD44), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were measured using a LightCycler® 96 (Roche Diagnostics, Mannheim, Germany). The genes were ampli ed using SYBR Green PCR SuperMix (SYBR High-Sensitivity qPCR SuperMix, Novoprotein, Shanghai, China) and 300 nM of each primer pair. Primers were designed using the Primer Design-online Software (Sangon Biotech, Shanghai, China) and synthesized by Sangon Biotech (Shanghai, China). Normalized gene expression levels are provided as the ratio between the mean value for the target gene and that for the reference gene (GAPDH) in each sample. All reactions were repeated three times. All RNA samples exhibited a 260/280 ratio 1.8. The primer sequences used are shown in Table 1.   Expression of NRF2 in PBMCs of patients with silicosis using immuno uorescence After PBMCs were isolated from peripheral blood, the expression of NRF2 in PBMCs of patients with silicosis was determined using a uorescent immunocytochemistry assay. PBMCs were diluted to 2×10 6 /ml with phosphate-buffered saline (PBS). A total of 300 µl of diluted suspension was spread evenly over the slide and allowed to air-dry. Then, PBMCs were xed with 4% paraformaldehyde for 20 min at room temperature and washed three times with PBS. Sections were incubated with primary antibodies to NRF2 (Abcam, 1:100) overnight at 4°C and then incubated with Cy3-conjugated secondary antibodies (1:200) at room temperature for 50 min in the dark. Nuclei were labeled using 4', 6-diamidine-2'-phenylindole dihydrochloride and the results observed under a confocal laser scanning microscope (TCS STED-CW, Leica Microsystems, Mannheim, Germany). Immuno uorescence images were analyzed using Image J Software. Sixteen areas per sample covering the whole spectrum of blood smear, were evaluated in a blinded fashion. The positive cells showed red uorescence in the nucleus. Finally, the Nrf2 positive cell rate in PBMCs was calculated.
Expression of NRF2 and NRF2 regulated proteins in PBMCs of patients with silicosis using western blotting Total proteins were extracted from PBMCs isolated from peripheral blood. Protein concentration was quanti ed using a BCA assay kit (Thermo, USA). Equal amounts of total protein were separated by 10% sodium dodecyl sulfatepolyacrylamide gel electrophoresis and then transferred onto a polyvinylidene uoride membrane (Millipore). After blocked for 2 h in 5% nonfat milk, they were incubated with the

Statistics
Routine test data were analyzed using SPSS 19.0 analysis software. Data are reported as the mean ± standard error of the mean. Group differences were calculated using a t-test and a Chi-squared test. RNA sequencing data between two groups were analyzed using the R statistical software package. P < 0.05 was considered to be signi cantly different.

Results
Clinical and cytology characteristics of patients with silicosis and healthy volunteers A total of 92 male patients with silicosis and 87 male healthy volunteers participated in this study; among them, only 77 patients with silicosis and 62 healthy controls fully satis ed the enrollment criteria.
Clinical and peripheral blood cytology characteristics of the subjects are described in Table 2. There were no statistical differences in age, BMI, blood pressure, and smoking status between the two groups. FVC and FEV1 for respiratory function were signi cantly decreased in the group of patients with silicosis. The number of monocytes increased signi cantly through cytology detection in patients with silicosis, indicating that monocytes may play an important role in cellular responses induced in patients with silicosis ( Table 2). Expression of NRF2 and NRF2-dependent antioxidative genes in PBMCs from patients with silicosis To investigate the potential role of NRF2 in PBMCs from patients with silicosis, the expression of NRF2 was examined using a confocal laser scanning microscope. Notably, the expression of NRF2 was signi cantly higher in patients with silicosis than in healthy controls (Fig. 1A). Semi-quantitative analysis revealed a signi cantly higher NRF2 fraction expressing positive cells in patients with silicosis compared to that in healthy controls (Fig. 1B). Moreover, RT-qPCR analysis showed that the expression of NRF2, KEAP1 and GCLM mRNA in PBMCs from patients with silicosis was increased compared to that in healthy controls (Fig. 1C). Accordingly, the protein expression in PBMCs from patients with silicosis were measured by western boltting. The results of western bolt showed the protein expression of NRF2, KEAP1 and GCLM in PBMCs increased in patients with silicosis compared with healthy controls (Fig. 1D and 1E). Moreover, to assure the process of epithelial mesenchymal transition in silicosis. These results showed that the expression levels of NRF2 and NRF2-dependent antioxidative genes increased in PBMCs from patients with silicosis. Network analysis of the DEGs from transcriptome sequencing found that the genes in the NRF2 module related to NRF2 involved in the development of silicosis, which suggests that DEGs associated with NRF2 might affect the pathogenesis of silicosis (Fig. 1F).

Identi cation of DEGs in the PBMCs of patients with silicosis
In total, six samples were included: three samples in the healthy control group and three samples in the group of patients with silicosis. Using commercial RNA preparation kits, total RNA was prepared from PBMC samples and the quality of total RNA was con rmed using the Agilent 2100 RNA 6000 Nano kit.
To identify key pathways and genes in the progression of silicosis, the differentially expressed mRNAs among healthy controls and patients with silicosis were analyzed using the Gene-e software. Three conditions were set for DEG screening: a false discovery rate < 20%, unfolded change > 2 and P < 0.05. A total of 1158 genes were identi ed after the analysis, of which, 475 were upregulated and 683 were downregulated ( Fig. 2A and 2B).
Key molecules and signaling pathways were screened through annotation of DEGs in the development of silicosis GO analysis was performed to explore potential roles of differentially expressed mRNAs in the development of silicosis. GO analysis showed that the DEGs between patients with healthy controls and patients with silicosis were mainly involved in cellular component (including organelle, extracellular region, extracellular matrix, and extracellular region part), oxidation reduction (such as catalytic and oxidoreductase activity), transcription factor activity (such as transcription factor activity, protein binding, and nucleic acid binding transcription factor activity), regulating metabolic processes, immune system processes, response to stimulus transcription, regulating signal transduction, and biological adhesion (Fig. 2C ).
KEGG pathway analysis showed that the DEGs between patients with silicosis and healthy controls were primarily enriched in pathways associated with regulating immune-related pathways (Graft-versus-host disease, Rheumatoid arthritis, Allograft rejection, Systemic lupus erythematosus, In ammatory bowel disease, Asthma, Human T-cell lymphotropic virus type 1 infection, Th1 and Th2 cell differentiation, and Th17 cell differentiation), regulating metabolic processes (amino acid metabolism, lipid metabolism, and metabolism of cofactors and vitamins), cell growth and death, cellular community, signal transduction (including the cGMP-dependent protein kinase G signaling pathway and the TGF-β signaling pathway), and transport and catabolism processes (cell adhesion molecules signaling pathway and the phagosome signaling pathway) (Fig. 2D).

NRF2-regulated key DEGs in PBMCs from patients with silicosis
Based on the above results, NRF2 played a key role in silicosis. GSEA, a powerful tool to infer biological function, was performed and showed that genes associated with GSH metabolism, the TGF-β signaling pathway, and ECM receptor interaction signaling pathway, which were closely related to silicosis, were signi cantly enriched in the group of patients with silicosis (Fig. 3A). These observations suggested that NRF2 may be closely related to the lymphocyte oxidative state in patients with silicosis.
We screened 27 DEGs that genes associated with GSH metabolism, the TGF-β signaling pathway, and ECM receptor interaction signaling pathway from 1158 DEGs (Table 3). To further de ne the interaction between the screened 27 DEGs, the STRING database was used to construct a PPI network. The PPI network consisted of 37 nodes interacting with 144 edges (Fig. 3B). The PPI database was ltered using a combined score > 0. 4   Legends RT-qPCR was performed to determine the expression of NRF2-regulated DEGs in PBMCs from patients with silicosis, and con rm selected RNA-Seq ndings. Steady-state mRNA levels of key genes associated with the TGF-β signaling pathway were determined. SMAD2, a downstream molecule of the TGF-β signaling pathway, was signi cantly increased in PBMCs from patients with silicosis. However, SMAD3 mRNA levels were signi cantly decreased compared to those in healthy controls. Next, it was con rmed that gene expression of BMP4 and MAPK3 were increased in PBMCs from patients with silicosis. Furthermore, the transcriptional factor related to cell-to-matrix interactions, THBS1, was signi cantly increased in PBMCs from patients with silicosis. Additionally, ITGB3 was increased and CD44 was signi cantly increased in PBMCs from patients with silicosis (Fig. 3C).

Discussion
Nuclear factor erythroid 2-related factor 2 (NRF2) is a transcription activator that regulates the expression of target genes by binding antioxidant response elements (AREs) [36]. NRF2 is important for the coordinated upregulation of genes in response to oxidative stress. Silica causes the signi cant accumulation of ROS and activates the antioxidative protein NRF2 and its downstream proteins in the early exposure to silica [37]. Immune cells with anti-in ammatory properties, especially monocytes/macrophages are involved in the possibility of alternative mechanisms of silicosis brosis [38][39][40]. The present study focused on the lymphocyte oxidative state and found that the number of monocytes and the expression of NRF2 and NRF2-dependent antioxidative genes was signi cantly increased in PBMCs from patients with silicosis. Additionally, it was con rmed that respiratory function was signi cantly decreased in the patients with silicosis.
The respiratory dysfunction in patients with silicosis is closely related to the severity of pulmonary brosis [41]. The TGF-β pathway can play an important direct inducer role in the process of collagen transcription in the development of silicosis [42], as TGF-βs are overexpressed in brosis [43]. TGF-β signaling plays a key role in extracellular matrix remodeling, the epithelial-mesenchymal transition, and cell growth, migration, differentiation in brosis [44]. Generally, TGF-β signaling is modulated by the phosphorylation of the cytoplasmic SMAD signaling molecules, which results in their translocation to the nucleus [45]. TGF-β1 signaling molecules play a key role by promoting transdifferentiation the broblast into myo broblasts, which promote collagen synthesis and ECM deposition in the pathology of silicosis [43].
The ECM, complex mixture of structural and functional macromolecules, has an important role in tissue brosis and in the maintenance of cell and tissue structure and function [46]. Integrins and other transmembrane molecules mediate speci c interactions between cells and the ECM [47,48]. These interactions have the direct and indirect effects of cellular activities, lead to adhesion, proliferation, apoptosis, migration, and differentiation. In addition, integrins act as 'mechanoreceptors' that they would provide a speci c physical link between the cytoskeleton and the ECM [48].
In the present study, the differential expression of mRNAs in the progression of silicosis was identi ed using RNA-Seq analysis. A total of 1158 dysregulated mRNAs were identi ed in PBMCs from patients with silicosis, including 475 upregulated and 683 downregulated mRNAs. GO analysis revealed that the functions of dysregulated mRNAs in PBMCs from patients with silicosis were related to the ECM, catalytic activity, oxidoreductase activity, transcription factor activity, metabolic processes, immune system processes, response to stimulus transcription, and biological adhesion. In KEGG pathway analysis, the dysregulated mRNAs were involved in regulating immune-related pathways, regulating metabolic processes, cellular community, cell adhesion molecules signaling pathway and the phagosome signaling pathway. Therefore, it is possible that the dysregulated mRNAs in these processes are involved in the pathogenesis of silicosis.
To con rm the results obtained by RNA-Seq analysis, seven differentially expressed mRNAs were selected to verify their expression in PBMCs from patients with silicosis using RT-qPCR. The results indicated that the expression of SMAD2, MAPK3, THBS1, SMAD3, ITGB3, and BMP4 was increased, while the expression of CD44 was signi cantly decreased in PBMCs from patients with silicosis compared to that of the control group, indicating consistent results with the RNA-Seq data. The TGF-β family of cytokines signals through receptor serine/threonine kinases to control cell behavior and fate [49]. These signals are propagated through the transcription factors SMAD2 and SMAD3 downstream of TGFβ. SMAD2 and SMAD3 belong to the SMAD protein family that mediate multiple signaling pathways as transcriptional modulators [50]. These ndings indicate that dysregulation of the TGFβ1-SMAD signaling pathway may play an important role in the pathological process of silicosis, and are consistent with ndings in animal experiments [51,52]. SMAD proteins mediate signaling of TGF-β through its interaction with the SMAD anchor for receptor activation (SARA) protein [53]. In response to a TGF-β signal, SMAD proteins are phosphorylated by TGF-β receptors [54]. This study revealed that BMP4 was a differential gene hub in PBMCs from patients with silicosis. BMP4 encodes a secreted ligand of the TGF-β superfamily of proteins and activates SMAD family transcription factors that regulate gene expression [55]. These results showed that SMAD2 and SMAD3 had different trends in the two groups. Recently, SMAD2 and SMAD3 have shown different roles in the TGF-β signaling pathway during embryonic development [50]. The cell experiments revealed that BMP7 is associated with inhibiting silica-induced brosis through activated BMP7/SMAD and suppressed TGF-β/SMAD pathways [56,57]. To date, BMP4 in the BMP protein family has not been reported in silicosis.
MAPK3 is especially involved in activation towards microtubule-associated protein-2 and the control of cell survival, proliferation and differentiation [58]. In the present study, MAPK3 (ERK1) was signi cantly increased in PBMCs from patients with silicosis compared to that in the control group. Dysregulation of MAPK3 plays a signi cant role in the pathological processes of silicosis [59]. Crystal compounds in silicosis activate ROS, which activate the in ammasome through MAPK3 [59,60]. In agreement with these ndings, activation of MAPK3 (ERK1) and NF-κB in PBMCs is reported during oxidative stress [61].
The transcription factor thrombospondin-1 (THBS1) is an adhesive glycoprotein that mediates cell-to-cell and cell-to-matrix interactions. THBS1 binds to cell surface receptors, including brinogen, bronectin, laminin, type V collagen and integrins, such as ITGB3 (Calzada and Roberts, 2005). These studies suggest that THBS1 in human peripheral blood lymphocytes is involved in the regulation of pesticideinduced immune dysfunction [62].
The protein encoded by CD44 is a cell-surface glycoprotein involved in cell-cell interactions, cell migration and adhesion [63]. It also interacts with other ligands, such as matrix metalloproteinases, and collagens [64]. CD44 blockade alleviates silica-induced brosis and improves pulmonary function in vivo [65].
Nonetheless, the present study remains unclear that changes speci c to different cell types present in PBMCs. Additionally, the effects of oxidative stress on the central signaling pathway, such as insulin signaling or glucose transport in silicosis patient should be addressed in subsequent investigations.

Conclusions
Genome-wide mRNA pro ling from PBMCs in patients with silicosis is identi ed. NRF2 signaling regulates the lymphocyte oxidative state and may contribute to brogenic responses in human PBMCs. Therefore, NRF2 might serve as a novel preventive and therapeutic candidate for silicosis. The novel information provided by this study contribute to understand of the oxidative stress mechanism of silicosis. Ethics approval and consent to participate All subjects signed an informed consent before being enrolled. The study was approved by the Ethics Committee of the Xinxiang Medical University (protocol number XYLL-2017086).

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.
Author details