Transcriptomics profile of human bronchial epithelial cells exposed to ambient fine particles and influenza virus (H3N2)

Fine particulate matter (PM2.5) pollution remains a major threat to public health. As the physical barrier against inhaled air pollutants, airway epithelium is a primary target for PM2.5 and influenza viruses, two major environmental insults. Recent studies have shown that PM2.5 and influenza viruses may interact to aggravate airway inflammation, an essential event in the pathogenesis of diverse pulmonary diseases. Airway epithelium plays a critical role in lung health and disorders. Thus far, the mechanisms for the interactive effect of PM2.5 and the influenza virus on gene transcription of airway epithelial cells have not been fully uncovered. In this present pilot study, the transcriptome sequencing approach was introduced to identify responsive genes following individual and co-exposure to PM2.5 and influenza A (H3N2) viruses in a human bronchial epithelial cell line (BEAS-2B). Enrichment analysis revealed the function of differentially expressed genes (DEGs). Specifically, the DEGs enriched in the xenobiotic metabolism by the cytochrome P450 pathway were linked to PM2.5 exposure. In contrast, the DEGs enriched in environmental information processing and human diseases, such as viral protein interaction with cytokines and cytokine receptors and epithelial cell signaling in bacterial infection, were significantly related to H3N2 exposure. Meanwhile, co-exposure to PM2.5 and H3N2 affected G protein-coupled receptors on the cell surface. Thus, the results from this study provides insights into PM2.5- and influenza virus-induced airway inflammation and potential mechanisms.


Influence of PM 2.5 on infectivity of H3N2 virus on BEAS-2B cells
To assess whether exposure to PM 2.5 could affect the ability of the H3N2 virus to infiltrate BEAS-2B, we first exposed them to PM 2.5 (12.5 g/mL) for 4 h before infecting them with H3N2 (MOI = 1).As shown in Fig. 3A  and B, co-exposure of BEAS-2B cells to PM 2.5 and H3N2 virus increased 65.9% of virus-infected cells compared to H3N2 group alone (P < 0.05).It suggests that exposure to PM 2.5 can increase the susceptibility of respiratory epithelial cells to IAV in co-exposure.This is consistent with what we found in our previous study 28 .

Analyses of gene expression and correlations
The distribution of the gene expression data for all samples is displayed in boxplots.Gene expression levels of all samples were essentially the same after normalization, indicating that batch effect and systematic bias were not significant (Fig. 4A).The results of the principal component analysis (PCA) revealed that the control (MOCK) group, H3N2 group, PM 2.5 group, and the co-exposure group (PM 2.5 + H3N2) were well separated, indicating that there was a high degree of similarity among the samples in the same group and differences among different groups (Fig. 4B).As shown in Fig. 4C, the correlation coefficient between the samples was calculated by Pearson correlation analysis, and the closer the correlation coefficient is to 1, the higher the similarity between the samples, and the smaller the differences between the samples.The results show that biological experimental operations of the samples are highly repeatable and that samples in the same experimental group have a high degree of similarity.www.nature.com/scientificreports/

Screen of DEGs
In PM 2.5 group, 53 DEGs were detected in contrast to the control group, of which 30 were up-regulated and 23 were down-regulated.In H3N2 group, 54 DEGs were detected with 21 up-regulated and 33 down-regulated.
In the co-exposure group, 97 DEGs were discovered compared with the control group, of which 45 were upregulated and 52 down-regulated.Moreover, 52 DEGs were found between the co-exposure group and PM 2.5 group, of which 22 were up-regulated and 30 down-regulated.47 DEGs were found between the co-exposure group and H3N2 groups, of which 32 were up-regulated and 15 down-regulated (Fig. 5A-C).These DEGs could be divided into multiple sub-groups through a heat map of hierarchical cluster analysis (Fig. 5D-F).The top 10 DEGs in each comparison group are displayed in Tables 1, 2, 3.

GO analysis of DEGs
The DEGs were further analyzed by GO enrichment analysis for their expression and functions.Hypergeometric test between the co-exposure group and the control group revealed 251 significant enrichment sub-classes related to the biological process (73.71%), molecular function (17.53%), and cellular components (7.37%), such as extracellular matrix organization, cellular protein metabolism, G protein-coupled receptor signaling pathway, and peptide hormone binding (Fig. 6A).
Between the co-exposure group and PM 2.5 group, 103 enrichment sub-classes were identified, which were related to the biological process (59.22%), molecular function (25.24%), and cellular component (15.54%), such as transmembrane signaling receptor activity, activation of GTPase activity, positive regulation of cell migration, intracellular protein transport, and plasma membrane (Fig. 6B).
Between the co-exposure group and H3N2 group, 119 enrichment sub-classes were identified, which were associated with the biological process accounting (68.07%), cellular component (8.10%), and molecular function  www.nature.com/scientificreports/(23.53%), such as extracellular matrix organization, cellular protein metabolism, G protein-coupled receptor signaling pathway, and peptide hormone binding (Fig. 6C).

KEGG pathway analysis of DEGs
Analysis of DEG-associated pathways was performed using the KEGG database to study the potential mechanism of the joint biological effects of PM 2.5 exposure and H3N2 infection.It was revealed that biological systems related to DEGs between the co-exposure and the control groups included the endocrine system, digestive system, and immune system.Related metabolic processes including lipid metabolism, carbon metabolism, etc. DEG-related human diseases include viral, bacterial, and immune diseases.DEG-related environmental information processing included signal transduction, signal molecules and interaction, and cellular community-eukaryotes of cellular processes (Fig. 7A).Moreover, DEGs were significantly enriched in several related pathways (levle3), including the calcium signaling pathway, neuroactive ligand-receptor interaction, complement and coagulation cascades, long-term potentiation, glucagon signaling pathway, and chemical carcinogenesis, among others.(Fig. 7B).
The biological systems associated with DEGs between the co-exposure and PM 2.5 groups included the immune system and endocrine system.DEG-associated human diseases included viral, bacterial, immune, and cancer.DEG-associated environmental information processing included signal transduction, signal molecules and interaction (Fig. 7C).Moreover, DEGs were significantly enriched in several related pathways (levle3), including the cytokine-cytokine receptor interaction, viral protein interaction with cytokine and cytokine receptors and epithelial cell signaling in Helicobacter pylori infection (Fig. 7D).
Biological systems associated with DEGs between the co-exposure and H3N2 groups included the immune system and the endocrine system.DEG-associated human diseases included viral, bacterial, immune, and cancer.In addition, DEG-associated environmental information processing included signal transduction, and signal molecules and interaction (Fig. 7E).DEGs were significantly enriched in several related pathways (levle3), including neuroactive ligand-receptor interaction, metabolism of xenobiotics by cytochrome P450, complement and coagulation cascades, chemical carcinogenesis, rheumatoid arthritis, and systemic lupus erythematosus, et al. (Fig. 7F).It is noteworthy that some differential genes and pathways were identified and enriched only in the coexposure group through the discovery of DEGs and KEGG enrichment, KEGG pathways such as calcium signaling pathway, long-term potentiation, etc., and related DEGs including Adrenoceptor Alpha 1 Beta (ADRA1B), Adrenoceptor Beta 1 (ADRB1), Calcium/Calmodulin Dependent Protein Kinase II Beta (CAMK2B), etc., implying that the co-exposure of PM 2.5 and H3N2 is more than just an additive effect.

Gene set enrichment analysis (GESA) of DEGs
The samples were split into two groups in the gene expression matrix that served as the input for the GSEA analysis.Then, the genes in these two groups were all listed based on their values generated from GSEA analysis from large to small, respectively.Here, we can tell whether a pathway is activated or inhibited by simply looking  www.nature.com/scientificreports/ at the value of its corresponding gene in the list.The gene with a large value was putatively considered as upregulated.On the contrary, the gene with a smaller value was considered down-regulated.
Compared with the control group, the up-regulated pathways in the co-exposure group included drug metabolism cytochrome P450, DNA replication, metabolism of xenobiotics by cytochrome P450, small cell lung cancer, base excision repair, autoimmune thyroid disease, antigen processing and presentation, and type I diabetes mellitus, the down-regulated pathways include steroid biosynthesis, biosynthesis of unsaturated fatty acids, and cytokine-cytokine receptor interaction (Fig. 8A).
Compared with PM 2.5 group, the up-regulated pathways in the co-exposure group included the chemokine signaling pathway, antigen processing and presentation, the down-regulated pathways include base excision repair, steroid biosynthesis, circadian rhythm mammal, insulin signaling pathway, JAK-STAT signaling pathway, and peroxisome (Fig. 8B).
Compared with H3N2 group, the upregulated pathways in the co-exposure group included metabolism of xenobiotics by cytochrome P450, drug metabolism by cytochrome P450, antigen processing and presentation, and TGF -beta signaling pathway, the down-regulated pathways include epithelial cell signaling in helicobacter pylori infection, tight junction, and pathogenic Escherichia coli infection (Fig. 8C).

Verification of DEGs using RT-PCR
The expression of some representative DEGs identified by transcriptomic methods within different groups was verified using RT-PCR.It was shown that mRNA levels of some DEGs including CYP1A1, CYPIB1 and Aldehyde Dehydrogenase 3 Family Member A1 (ALDH3A1) in the co-exposure group were significantly higher than those in the control group and H3N2 group.Such results provide clues for exploring the mode of action of PM 2.5 in combined exposures.
The mRNA level of Interleukin 21 Receptor (IL21R) was significantly higher in the PM 2.5 exposure group and significantly lower in the H3N2 and co-exposure groups.Furthermore, the mRNA level of ADRB1 was significantly lower in the co-exposure group, but not in the exposure alone group (P < 0.05), and CAMK2B was similar, with a tendency of lower expression in the co-exposure group.(Fig. 9).These findings were in support of the results from RNA-seq, as depicted above.

Discussion
Even though PM 2.5 air pollution has been significantly reduced in recent years, it remains a major threat to public health.A previous study has reported that PM 2.5 has an impact on the infection and severity of infectious diseases 29 .In this study, we examined the joint effect of PM 2.5 and H3N2 on gene transcription in human bronchial epithelial cells using RNA-seq.The results suggest that pleiotropic genes and pathways are involved in the promotive effect of PM 2.5 on H3N2 infection of human bronchial epithelial cells.
To explore the effect of PM 2.5 exposure on H3N2 infection of BEAS-2B cells and the possible underlying mechanisms, we first examined the differential effect of the co-exposure group and PM 2.5 group on gene expression, 22 DEGs are found up-regulated and 30 DEGs down-regulated between these two groups.Analysis of the related functions of these DEGs suggest H3N2 exposure may facilitate viral protein interaction with cytokine and cytokine receptor, and epithelial cell signaling in Helicobacter pylori infection in the co-exposure group are significantly enriched pathways.Whether H3N2 exposure modulates the effect of PM 2.5 on BEAS-2B cells through signaling molecules and interaction of environmental information processing and bacterial infectious diseases is currently under the assumption.
Previous studies have reported the interaction between viral and bacterial infections.TLR4 is proposed as unable to recognize viruses, but RSV infection has been shown to upregulate TLR4 expression, which increases inflammatory signaling and makes the respiratory system more sensitive to LPS, the major surface membrane component present in almost all Gram-negative bacteria and also a common biological component absorbed on PM 2.5 . 30Influenza infection followed by secondary bacterial pneumonia is associated with significant mortality and mortality.The susceptibility to bacterial infection may be increased by viral pathogen-associated molecular  patterns (PAMP) desensitization to TLRs.Desensitization results in reduced chemokine production and NF-κB activation 31 .The hygiene hypothesis for asthma pathogenesis is also based on the state of tolerance after repeated PAMP exposure, early exposure to viral PAMP may lessen the risk of developing high inflammation later in life 32 .In another of our previous experiments 28 , the effects of PM 2.5 exposure on influenza virus (H3N2) infection and downstream regulation of inflammatory and antiviral immune responses were investigated, also using the human bronchial epithelial cell line BEAS-2B.The results showed that exposure to PM 2.5 alone increased the production of pro-inflammatory cytokines, including interleukin-6 (IL-6) and IL-8, but decreased the production of the antiviral cytokine interferon-β (IFN-β) in BEAS-2B cells.In contrast, exposure to H3N2 alone increased the production of IL-6, IL-8, and IFN-β.In this study, PM 2.5 exposure significantly increased IL21R expression, whereas IL21R expression was decreased in both groups with H3N2 exposure.A recent study has shown that IL-21R signaling suppresses IL-17 + gamma delta T cell responses and production of IL-17 related cytokines in the lung at steady state and after influenza A virus infection 33 .These findings suggest that H3N2 may influence the effects on cells via signaling molecules and environmental information processing with bacterial infectious diseases interactions of viral proteins with cytokines and cytokine receptors, and epithelial cell signaling in combined exposures.
Meanwhile, we identified 32 up-regulated and 15 down-regulated DEGs between the co-exposure group and the H3N2 group.The metabolism of xenobiotics by cytochrome P450, complement and coagulation cascades, and chemical carcinogenesis are significantly enriched pathways associated with these DEGs.Whether PM 2.5 exposure modulates the response of BEAS-2B cells to H3N2 through the pathways regulated by cytochrome P450 metabolism, respiratory immune system, or cancer-related diseases, needs to be clarified in the future.
Moreover, the expression of some DEGs in BEAS-2B cells exposed to PM 2.5 and H3N2 are examined using RT-PCR.Similar to the results from RNA-seq, mRNA levels of CYP1A1 and ALDH3A1 in the co-exposure group, those of CYP1B1 and ALDH3A1 H3N2 group are significantly higher than those in the control group, respectively.KEGG indicates that the pathway related to ALDH3A1 is associated with the metabolism of xenobiotics by cytochrome P450 and chemical carcinogenesis.Additionally, differential expression of CYP1A1 and CYP1B1 is detected between the co-exposure group and control group or H3N2 group, respectively.The CYP1 (cytochrome P450 1) family has two significant subtypes, CYP1A1 and CYP1B1, which are abundant in lung tissues.CYP1A1 and CYP1B1 participate in the metabolism of lung polycyclic aromatic hydrocarbons (PAHs) as PAHs-sensitive genes and can be activated by PAHs in the lung through the aryl hydrocarbon receptor (AhR) [34][35][36] .CYP1A1 is most abundant in alveolar type II cells and endothelial cells 37 , while CYP1B1 is most abundant in airway epithelial cells 35 .Interestingly, PAH content contributes to PM 2.5 toxicity, and recent research has shown that PAHs in PM are a powerful mediator of health effects [38][39][40][41] .Mice exposed to high PM concentrations in Fresno, California were found to increase CYP1A1 expression in pulmonary tissues, including pulmonary blood vessels, parenchyma tissue, and airways 42 .Another study found that CYP1B1, as an enzyme with a unique tumor-specific expression pattern, can bioactivate a wide range of carcinogenic compounds.Inflammatory cytokines such as tumor necrosis factor-α (TNF-α) and AhR ligands co-regulate CYP1B1 expression and change the metabolism of exogenous carcinogens, increase the biological activity of promutagens, such as benzo[a]pyrene (BaP) in epithelial cells 43 .Recent studies have shown that CYP450 is able to influence macrophage inflammatory signaling through the PPARa axis, which may explain how PM 2.5 affects the role of H3N2 viruses in co-exposure 44 .These findings imply that the action pattern of PM 2.5 may be closely related to the metabolism of cytochrome P450 in the context of co-exposure to H3N2 and PM 2.5 .
Notably, this study discovered that some differential genes and pathways were identified and enriched only in the common exposure group, KEGG pathways such as calcium signaling pathway, long-term potentiation, and related DEGs including ADRA1B, ADRB1, CAMK2B, and so on, and the results of RT-PCR assay confirmed the DEGs differences among them.ADRA1B and ADRB1 belong to the G protein-coupled receptor adrenergic receptor group and GO annotations for this gene include G protein-coupled receptor activity α1 adrenergic receptor activity, and β adrenergic receptor activity.Calmodulin-dependent kinases (CaMK) are a family of serine/ www.nature.com/scientificreports/threonine kinases that mediate many of the second messenger effects of Ca 2+ .Recent studies have demonstrated that CAMK2B expression is modified in neuropsychiatric illnesses and potentially affects synaptic plasticity 45 .These findings suggest that co-exposure to PM 2.5 and H3N2 may be able to affect the calcium-sensing receptor on the cell surface by modulating Ca 2+ , which, as a typical nutrient-sensing G-protein-coupled receptor, is activated and modulated by a wide range of endogenous or exogenous substances (e.g., cations, amino acids, polyamines, aminoglycoside antibiotics, etc.), resulting in an "additional" effect 46 .Its specific mechanisms will be investigated further.
In conclusion, PM and H3N2 still pose serious risks to public health, but the mechanisms of their combined action remain largely unknown.This study profiles the transcriptome of human bronchial epithelial cells exposed to PM 2.5 influenza virus (H3N2) by RNA-Seq.The results indicate that PM 2.5 exposure disrupts the expression of CYP-coding genes, further altering the body's metabolism of exogenous harmful substances, and leading to the intensification of H3N2 invasion of BEAS-2B cells.Meanwhile, by working as a PAMP, H3N2 exposure might influence the immune system's response to PM 2.5 that contains bacterial pathogens or LPS.We also found that the combined effects of PM 2.5 and H3N2 are not simply additive and that the combined exposure of the two may have an "additional" effect by modulating Ca 2+ to affect G protein-coupled receptors on the cell surface.The exploration of the joint effects of PM 2.5 and H3N2 may provide insights into the pathophysiological basis of the interaction between PM and influenza virus, as well as for developing efficient strategies to prevent the adverse respiratory effects caused by PM 2.5 and H3N2 viruses.

Materials and methods PM 2.5 collection and suspension preparation
Quartz sampling filters were used to collect PM 2.5 in January and March 2021 using a TischTE-6070 high-flow particle sampler (Tisch Environmental, USA) with a flow rate of 1.13 m 3 /min.After sample collection, the quartz sampling filter was submerged in a Petri dish 10 cm in diameter filled with ultrapure water and sonicated three times for five minutes each.The suspension was filtered through six layers of gauze and then underwent lyophilization to collect PM 2.5 powder using a vacuum freeze drier (Christ, Germany) for 24 h.PM 2.5 was thoroughly suspended in phosphate-buffered saline (PBS, Solarbio Life Sciences, China) solution at a final concentration of 1 mg/mL.The suspension was aliquoted and stored in a -80 freezer.Before use, PM 2.5 suspension was vortexed for homogenization.

Immunoperoxidase monolayer cell assay (IPMA)
IPMA with 3-Amino-9-Ethylcarbazole (AEC) peroxidase substrate as the chromogenic solution was used to visualize H3N2-infected BEAS-2B cells.The 12-well plate was cleaned once with PBS before addition of 5% skimmed milk to each well, and the wells were then blocked in a thermostat at 37 °C for 1 h.After blocking, monoclonal antibodies against hemagglutinin of H3N2 were incubated at 37 °C for 1 h, and then washed five times with PBS containing 0.05% Tween-20 (PBST).Goat anti-mouse IgG-horseradish peroxidase (HRP) antibody (Beyotime, China) was added and incubated for another 1 h at 37 °C.AEC chromogenic solution (Affinity Biosciences, USA) was applied to identify the infected cells.The number of infected cells was counted under a microscope (Leica, Germany) which showing a brownish-red color after treatment with AEC chromogenic solution were observed and counted.The average number of H3N2-infected cells per microscopic field was determined according to a protocol for a systematic randomization procedure 47 , and 8 images were recorded for each experimental group, from 3 independent experiments.www.nature.com/scientificreports/

RNA isolation and library preparation
Total RNA was extracted using the miRNA Isolation Kit (mirVana™, Ambion-1561) according to the manufacturer's instructions.The NanoDrop 2000 spectrophotometer was used to assess the purity and quantity of RNA (Thermo Scientific, USA).The Agilent 2100 Bioanalyzer was used to evaluate the integrity of the RNA (Agilent Technologies, Santa Clara, CA, USA).The mRNA libraries were then created using the TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA) for carrying out the transcriptome sequencing and analysis (OE Biotech Co., Ltd., Shanghai, China).

Quality control and RNA sequencing
PCA was used to analyze gene expression data and used to measure the distance between the samples in order to identify sample similarities.Pearson correlation analysis was used to calculate the correlation coefficients between samples, and the correlation between samples represents the degree of similarity between samples, and the similarity of samples from various treatments or tissues in terms of expression levels.The correlation of biological duplicates can be used to not only examine the reproducibility of biological experimental manipulations but also to evaluate the dependability of differentially expressed genes and to aid in the screening of aberrant samples.On the Illumina HiSeq X Ten platform, the mRNA libraries were sequenced, and 150 bp paired-end reads were generated.The HiSeq X Ten System was specially created for population-scale whole-genome sequencing.Each HiSeq X System is capable of 30-fold or greater coverage of human genome sequencing.Trimmomatic 48 was utilized for the processing of raw data (raw reads).To get the clean reads, the low-quality reads and reads containing ploy-N were eliminated.The human genome (GRCh38) was then mapped using HISAT2 49 using the clean reads.FPKM 50 of each gene was calculated using Cufflinks 51 , and the read counts of each gene were obtained by HTSeqcount 52 .Differential expression analysis was performed using the DESeq (2012) R package.

Differential expression gene (DEG) analysis
Only genes with count mean values larger than 2 were kept for the subsequent study after the genes had initially been filtered based on the count's mean value.The DESeq software was used to normalize the counts of each sample gene (the BaseMean value was used to estimate the expression), the difference fold was calculated and the NB (negative binomial distribution test) was used to test the significance of differences.Finally, the differential protein-coding genes were screened based on the fold and significance test results.For screening differences, the default parameters were P < 0.05 and |Log 2 foldchange|> 1.The expression pattern of genes in different groups and samples was displayed using hierarchical cluster analysis of DEGs.

GO and KEGG enrichment analysis
R software was used to analyze the DEGs' Gene Ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment based on the hypergeometric distribution.GO enrichment and KEGG 53 pathway enrichment analysis of DEGs were performed respectively using R based on the hypergeometric distribution.Biological process, cellular composition, and molecular function, the three categories of GO functional annotations, were all covered.The pathway diagram was derived from the KEGG database.

Gene set enrichment analysis
For Gene set enrichment analysis (GSEA), the GSEA software (version 3.0) was used to divide the samples into two groups according to the presence or absence of PM 2.5 exposure and download the samples from Molecular Signatures Database (http:// www.gsea-msigdb.org/ gsea/ downl oads.jsp) 54 , the c2.cp.kegg.v7.4.symbols.gmtsubset was downloaded to evaluate relevant pathways and molecular mechanisms based on gene expression profiles and phenotypic groupings, setting a minimum gene set of 5 and a maximum gene set of 5000.One thousand resamples with P < 0.05 was considered statistically significant.

Real-time quantitative reverse transcription polymerase chain reaction (RT-PCR)
Each RT reaction contained 10 μL of 5 × TransScript All-in-one SuperMix for qPCR, 2 μL of 0.5 μg RNA, and 0.5 μL of gDNA Remover.In a GeneAmp ® PCR System 9700 (Applied Biosystems, USA), reactions were carried out for 15 min at 42 °C and 5 s at 85 °C.After being diluted 10 times in nuclease-free water, the 10 μL RT reaction mix was kept at -20 °C.
RT-PCR was carried out using the LightCycler ® 480 II Real-time PCR Instrument (Roche, Switzerland) and a 10 μL PCR reaction mixture that contained 1 μL of cDNA, 5 μL of 2 × PerfectStart™ Green qPCR SuperMix, 0.2 μL of forward primer, 0.2 μL of reverse primer, and 3.6 μL of nuclease-free water.In a 384-well optical plate (Roche, Switzerland), reactions were incubated for 30 s at 94 °C, then underwent 45 cycles of 5 s at 94 °C and 30 s at 60 °C.Each sample was run in triplicate for analysis.Melting curve analysis was performed following the PCR cycles to confirm the precise generation of the desired PCR product.mRNA sequences obtained from the NCBI database served as the basis for the design of primer sequences.The primer sequences are shown in Table 4.The expression levels of mRNAs were normalized to ACTB (β-actin) and calculated using the 2-ΔΔCt method.

Statistical analysis
SPSS 22.0 (IBM, USA) and GraphPad Prism8 (GraphPad Software, USA) statistical software was used to process the experimental data, analyze the variables, and create graphs.Data with a normal distribution were expressed as mean and standard deviation ( X ± SD).ANOVA was applied to compare the cell viability of each group, and the Games-Howell method was used to compare the groups pairwise.The t-test of two independent samples

Figure 4 .
Figure 4. Gene expression level, principal component analysis, and correlations among samples.(A) The distribution of expression data from all the samples.(B) Principal component analysis (PCA).(C) The correlations among samples were analyzed by Pearson correlation.

Figure 5 .
Figure 5. Volcanic map of DEGs expression profiles among (A) Co-exposure and MOCK.(B) Co-exposure and PM 2.5 exposure.(C) Co-exposure and H3N2 virus exposure.And hierarchical cluster analysis heat map of DEGs expression profiles among (D) Co-exposure and MOCK.(E) Co-exposure and PM 2.5 exposure.(F) Co-exposure and H3N2 virus exposure.(P < 0.05, |log 2 FC|> 1).

Figure 6 .
Figure 6.GO enrichment analysis of DEGs between (A) Co-exposure and MOCK.(B) Co-exposure and PM 2.5 exposure.(C) Co-exposure and H3N2 virus exposure (Top 30 GO term).

Figure 7 .
Figure 7. KEGG pathway classification (upper panel) and enrichment analysis (lower panel) of DEGs between Co-exposure and MOCK (A and B), Co-exposure and PM 2.5 exposure (C and D), and Co-exposure and H3N2 virus exposure (E and F).

Table 1 .
Top 10DEGs between co-exposure group and MOCK group.

Table 2 .
Top 10DEGs between co-exposure group and PM 2.5 exposure group.

Table 3 .
Top 10DEGs between co-exposure group and H3N2 exposure group.