Advances in Virology

Advances in Virology / 2016 / Article

Research Article | Open Access

Volume 2016 |Article ID 3605302 | 11 pages |

A Cross-Study Biomarker Signature of Human Bronchial Epithelial Cells Infected with Respiratory Syncytial Virus

Academic Editor: Jay C. Brown
Received31 Dec 2015
Accepted13 Apr 2016
Published04 May 2016


Respiratory syncytial virus (RSV) is a major cause of lower respiratory tract infections in children, elderly, and immunocompromised individuals. Despite of advances in diagnosis and treatment, biomarkers of RSV infection are still unclear. To understand the host response and propose signatures of RSV infection, previous studies evaluated the transcriptional profile of the human bronchial epithelial cell line—BEAS-2B—infected with different strains of this virus. However, the evolution of statistical methods and functional analysis together with the large amount of expression data provide opportunities to uncover novel biomarkers of inflammation and infections. In view of those facts publicly available microarray datasets from RSV-infected BEAS-2B cells were analyzed with linear model-based statistics and the platform for functional analysis InnateDB. The results from those analyses argue for the reevaluation of previously reported transcription patterns and biological pathways in BEAS-2B cell lines infected with RSV. Importantly, this study revealed a biosignature constituted by genes such as ABCC4, ARMC8, BCLAF1, EZH1, FAM118A, FAM208B, FUS, HSPH1, KAZN, MAP3K2, N6AMT1, PRMT2, S100PBP, SERPINA1, TLK2, ZNF322, and ZNF337 which should be considered in the development of new molecular diagnosis tools.

1. Introduction

Respiratory syncytial virus (RSV) is a major etiologic agent causing acute lower respiratory infections that can progress to bronchiolitis and pneumonia in children, elderly, and immunocompromised individuals [1, 2]. RSV outbreaks are influenced by virus diversity and evolution [3, 4], environmental factors [5], and host immunity [6].

The epithelium is the primary site for host-virus interface, where cells recognize pathogen-associated patterns on microbes through innate immunity receptors [7, 8]. Indeed, epithelial cells constitute an important line of defense against RSV and other airborne pathogens [9]. They form a physical barrier and produce mucus to inhibit microbes from entering the body. Moreover, they express molecules with antimicrobial properties, as lysozyme, lactoferrin, collectins, and antimicrobial peptides [10]. Two human cell lines have been extensively used to understand the interaction between host and RSV, the alveolar epithelial cell, A549, and one from proximal airways, the bronchial epithelial cell, BEAS-2B.

Genome-wide microarrays are powerful tools to investigate host transcriptional response during infections in the pulmonary epithelium, including those induced by RSV [11, 12]. Indeed, two studies evaluated the patterns of gene expression from BEAS-2B cell lines infected with RSV [10, 13]. However, it is intriguing that after 4 h of infection Huang and collaborators (2008) found that RSV-modulated genes were only associated with the neuroactive ligand-receptor interaction pathway [13]; in contrast, Mayer and collaborators (2007) identified that the same time of RSV infection of BEAS-2B cells induced transcriptional changes similar to those found for other respiratory pathogens as Pseudomonas aeruginosa [10]. In spite of differences, publicly available microarray data offers an interesting opportunity to reveal common features of RSV induced transcriptional profiles to understand the early response of BEAS-2B cell lines and extend the knowledge on biomarkers of acute infections with this virus. Therefore, those datasets were evaluated in a meta-analysis by fitting linear models for each array probe and Empirical Bayesian approach to detect transcriptional changes that revealed significant associations with unreported pathways. Of importance, this strategy also rendered a biomarker signature of BEAS-2B cell lines infected with RSV that can be useful for the design of molecular diagnosis tools.

2. Materials and Methods

The datasets GSE3397 and GSE6802 were obtained from GEO database (, which compared BEAS-2B cells infected with RSV with control experiments. Only arrays in which cells were infected with RSV for 4 h were selected for further analysis. Raw data were processed using the R Language and Environment for Statistical Computing (R) 3.2.0 [14] and Bioconductor 3.1 [15]. The affy package for R [16] was used to perform quality control when applicable. Data was transformed and quantile normalization was applied for dataset GSE3397 due the absence of CEL files. The dataset GSE6802 was already RMA normalized. Batch effects were corrected with Combat( ) function [17] of sva package for R [18]. Expression data were weighted with the arrayWeights( ) function from limma package for R [19]. Differential gene expression was also evaluated with limma package for R [19], whereby differentially expressed genes (DEGs) were identified by a false discovery rate (FDR) <0.05. Hierarchical clustering was performed with Euclidian distance for metric calculations and the complete linkage method, which were displayed as heatmaps drawn with gplots package for R [20]. Pathway analyses were performed with the online platform for functional analysis InnateDB [21] and significant pathway overrepresentation was computed with hypergeometrical distribution and Benjamini-Hochberg correction for multiple comparisons. Significantly enriched pathways were determined by a value < 0.05 and FDR < 0.1.

3. Results and Discussion

3.1. Dataset Selection and Preprocessing Analysis

To define a robust transcriptional signature of BEAS-2B acutely infected with RSV, two publicly available datasets, GSE3397 and GSE6802, were used to conduct a meta-analysis from which data were extracted for BEAS-2B cells infected with RSV for 4 h and controls. First, background subtracted expression data from GSE3397 (Figure 1(a)) were preprocessed and normalized (Figure 1(b)). However, in a first attempt to conduct differential gene expression analysis using limma [19], there were no statistically significant differences in gene expression. Therefore, principal component analysis (PCA) was used to evaluate the expression profiles of each array and, except for arrays named here Control2 and RSV2, the consistent pattern of clustering in Figure 1(c) suggests a batch effect. After normalization, this effect was even more evident (Figure 1(d)), which led to the speculation that Huang and collaborators (2008) [13] analyzed only three microarray experiments from this dataset based on the assumption that differences found for those microarrays were due to failures in experimental procedures; however they did not consider or correct for batch effects. In view of those facts, the datasets were adjusted with Combat function for R, which removed such effects from GSE3397 expression data (Figure 1(e)). Batch correction of GSE3397 did not change the profiles of arrays Control2 and RSV2; nevertheless, those arrays were included in further analysis because the variation observed in this experiment could have a substantial impact over the final result. Even adverse experimental variations that may change the overall expression patterns of a dataset could be useful to power up the identification of genes that are robustly modulated in BEAS-2B cells infected with RSV. The expression dataset GSE6802 (Figure 1(f)) was also included in the analysis. PCA from expression data extracted from GEO demonstrates that most of the variability between the arrays is explained (76.6%) by the infection with RSV, as the standardized PC1 separates RSV-infected from control arrays (Figure 1(g)), whereas standardized PC2 (11.4%) separates one pair of arrays (RSV_3 and ctrl2) and, although these arrays are supposedly from different batches, clustering features of this axis also suggested a batch effect (Figure 1(g)). log2 transformation of data impacted the profile of array RSV_1 however did not change the profiles from RSV_3 and ctrl_2 (Figure 1(h)). Combat( ) function was also applied to the expression dataset GSE6802; however, PCA shows that the adjustment did not to improve further clustering between specific arrays (Supplementary Figure  1; see Supplementary Material available online at In view of that, downstream analyses were carried out with normalized log2 transformed data.

3.2. Differential Gene Expression

Next, linear model-based statistical analyses with a FDR < 0.05 were conducted to identify differentially expressed genes (DEGs). The dataset GSE3397 exhibited ninety-four DEGs (Figure 2(a) and Table 1). Those genes are highly discordant from DEGs previously reported by Huang and collaborators (2008) [13], which identified 277 DEGs based on different statistical analysis and assumptions. Fifty genes were downregulated and forty-four were upregulated (Table 1). The differences found in this study might reflect the inclusion of all microarray experiments from controls and 4 h after RSV infection; exclusion of expression data from 24 h after RSV infection; distinct preprocessing approaches as normalizing method and batch effect correction; and the assessment of statistical significance with a linear model-based method and corrected values. In contrast, 1965 DEGs were identified for the dataset GSE6802. The top hundred DEGs ranked by fold changes (Figure 2(b) and Table 2) included genes such as JUNB, KLF4, CXCL1, CXCL2, and IL6, which are in agreement with those reported by Mayer and collaborators (2007) [10]. Several factors should account for the notable differences in expression analysis from both datasets. First, different RSV strains were used to stimulate BEAS-2B cells. Second, experimental conditions of controls were also different, as control experiments from GSE3397 were incubated with vehicle (not specified) and those from GSE6802 were not stimulated. Third, despite both datasets being generated with affymetrix microarray platform, those include distinct versions, HU133 plus 2.0 for GSE3397 and HU133A 2.0 for GSE6802.

ProbeIDGene symbolGene name fold changeFDR

1560754_atCMTM7CKLF like MARVEL transmembrane domain containing 7−1,547560,017104
239439_atAFF4AF4/FMR2 family member 4−1,535810,023832
238929_atSRSF8Serine/arginine-rich splicing factor 8−1,518870,018433
223142_s_atUCK1Uridine-cytidine kinase 1−1,479390,017104
228007_atCEP85LCentrosomal protein 85 kDa-like−1,41030,017104
235573_atHSPH1Heat shock protein family H (Hsp110) member 1−1,399590,0371
228391_atCYP4V2Cytochrome P450 family 4 subfamily V member 2−1,387990,01671
219376_atZNF322Zinc finger protein 322−1,34910,046761
1553689_s_atMETTL6Methyltransferase like 6−1,347230,017104
242837_atSRSF4Serine/arginine-rich splicing factor 4−1,340710,044693
237215_s_atTFRCTransferrin receptor−1,326850,017104
208819_atRAB8ARAB8A, member RAS oncogene family−1,325930,042264
236665_atCCDC18Coiled-coil domain containing 18−1,314940,034201
206147_x_atSCML2Sex comb on midleg-like 2 (Drosophila)−1,305860,016454
229325_atZZZ3Zinc finger ZZ-type containing 3−1,304950,017104
1565716_atFUSFUS RNA binding protein−1,294150,049505
205062_x_atARID4AAT-rich interaction domain 4A−1,288770,033039
1552312_a_atMFAP3Microfibrillar associated protein 3−1,285210,046511
223223_atARV1ARV1 homolog, fatty acid homeostasis modulator−1,279870,023832
232001_atPRKCQ-AS1PRKCQ antisense RNA 1−1,279870,035983
233195_atDNAI1Dynein axonemal intermediate chain 1−1,259630,047083
219094_atARMC8Armadillo repeat containing 8−1,255270,043392
235232_atGMEB1Glucocorticoid modulatory element binding protein 1−1,24920,046511
218643_s_atCRIPTCXXC repeat containing interactor of PDZ3 domain−1,242290,0371
1566851_atTRIM42Tripartite motif containing 42−1,240570,042149
221821_s_atKANSL2KAT8 regulatory NSL complex subunit 2−1,237990,017104
244115_atFAM126AFamily with sequence similarity 126 member A−1,231140,033039
215541_s_atDIAPH1Diaphanous related formin 1−1,227740,033039
203196_atABCC4ATP binding cassette subfamily C member 4−1,225190,033039
225024_atRPRD1BRegulation of nuclear pre-mRNA domain containing 1B−1,222640,043765
37860_atZNF337Zinc finger protein 337−1,220950,023832
212997_s_atTLK2Tousled like kinase 2−1,218410,04814
225690_atCDK12Cyclin-dependent kinase 12−1,210830,0371
232103_atBPNT13′(2′), 5′-Bisphosphate nucleotidase 1−1,207480,0371
224848_atCDK6Cyclin-dependent kinase 6−1,202470,0371
214962_s_atNUP160Nucleoporin 160 kDa−1,202470,046319
219629_atFAM118AFamily with sequence similarity 118 member A−1,198310,028374
212290_atSLC7A1Solute carrier family 7 member 1−1,197480,042264
227187_atCBLL1Cbl proto-oncogene like 1, E3 ubiquitin protein ligase−1,195820,030047
233208_x_atCPSF2Cleavage and polyadenylation specific factor 2−1,193340,046319
230566_atMORC2-AS1MORC2 antisense RNA 1−1,176910,0371
238795_atFAM208BFamily with sequence similarity 208 member B−1,176090,0371
204980_atCLOCKClock circadian regulator−1,172830,0371
238653_atLRIG2Leucine-rich repeats and immunoglobulin like domains 2−1,172020,048527
229939_atENDOVEndonuclease V−1,168780,041349
218185_s_atARMC1Armadillo repeat containing 1−1,161510,046319
201083_s_atBCLAF1BCL2 associated transcription factor 1−1,155090,049505
227840_atC2orf76Chromosome 2 open reading frame 76−1,151090,042264
201686_x_atAPI5Apoptosis inhibitor 5−1,140760,046761
221699_s_atDDX50DEAD-box helicase 501,1407640,046511
1556178_x_atTAF8TATA-box binding protein associated factor 81,1590960,034358
205623_atALDH3A1Aldehyde dehydrogenase 3 family member A11,1639270,049505
212495_atKDM4BLysine demethylase 4B1,1933360,044693
1569057_s_atMIA3Melanoma inhibitory activity family member 31,1933360,047866
222494_atFOXN3Forkhead box N31,195820,048527
223311_s_atMTA3Metastasis associated 1 family member 31,195820,041439
215424_s_atSNW1SNW domain containing 11,1966490,049505
213478_atKAZNKazrin, periplakin interacting protein1,199140,025143
227864_s_atMVB12AMultivesicular body subunit 12A1,2016360,030287
228674_s_atEML4Echinoderm microtubule associated protein like 41,2041370,040345
224196_x_atDPH5Diphthamide biosynthesis 51,2058080,025143
224652_atCCNYCyclin Y1,2074810,046761
212968_atRFNGRFNG O-fucosylpeptide 3-beta-N-acetylglucosaminyltransferase1,2116730,0371
1555486_a_atPRR5LProline rich 5 like1,2125130,017104
232837_atKIF13AKinesin family member 13A1,2141950,042264
224320_s_atMCM8Minichromosome maintenance 8 homologous recombination repair factor1,2175660,033039
230131_x_atARSDArylsulfatase D1,2217930,0371
218225_atECSITECSIT signalling integrator1,2243360,034358
222610_s_atS100PBPS100P binding protein1,2268850,030047
32259_atEZH1Enhancer of zeste 1 polycomb repressive complex 2 subunit1,2294390,0371
203854_atCFIComplement factor I1,2328520,042264
221600_s_atAAMDCAdipogenesis associated, Mth938 domain containing1,2605030,0371
209558_s_atHIP1RHuntingtin interacting protein 1 related1,2631270,042264
224814_atDPP7Dipeptidyl peptidase 71,264880,016454
232280_atSLC25A29Solute carrier family 25 member 291,2772140,030047
228424_atNAALADL1N-Acetylated alpha-linked acidic dipeptidase-like 11,2869890,042264
203409_atDDB2Damage specific DNA binding protein 21,2887750,023832
229975_atBMPR1BBone morphogenetic protein receptor type 1B1,2977390,034358
227073_atMAP3K2Mitogen-activated protein kinase kinase kinase 21,2977390,017104
225347_atARL8AADP ribosylation factor like GTPase 8A1,2986390,02672
221774_x_atSUPT20HSPT20 homolog, SAGA complex component1,3085780,016454
223679_atCTNNB1Catenin beta 11,3185940,018043
227679_atHDAC11Histone deacetylase 111,3286860,044693
220020_atXPNPEP3X-Prolyl aminopeptidase 3, mitochondrial1,3425730,031097
203199_s_atMTRR5-Methyltetrahydrofolate-homocysteine methyltransferase reductase1,3603710,017104
228722_atPRMT2Protein arginine methyltransferase 21,3707830,016454
228951_atSLC38A7Solute carrier family 38 member 71,4319690,016454
217529_atORAI2ORAI calcium release-activated calcium modulator 21,4539730,043775
220311_atN6AMT1N-6 adenine-specific DNA methyltransferase 1 (putative)1,4600320,017104
213402_atZNF787Zinc finger protein 7871,4691690,017104
226055_atARRDC2Arrestin domain containing 21,4773380,017104
219756_s_atPOF1BPremature ovarian failure, 1B1,5800830,016454
202833_s_atSERPINA1Serpin peptidase inhibitor, clade A (alpha-1 antiproteinase, antitrypsin), and member 12,4880230,0371

ProbeIDGene symbolGene name fold changeFDR

212615_atCHD9Chromodomain helicase DNA binding protein 9−3,696090,00131
221840_atPTPREProtein tyrosine phosphatase, receptor type E−3,565240,000195
220817_atTRPC4Transient receptor potential cation channel subfamily C member 4−3,391680,001582
221703_atBRIP1BRCA1 interacting protein C-terminal helicase 1−2,887860,021463
207012_atMMP16Matrix metallopeptidase 16−2,826470,000119
219494_atRAD54BRAD54 homolog B (S. cerevisiae)−2,812790,000177
207034_s_atGLI2GLI family zinc finger 2−2,797230,005157
203518_atLYSTLysosomal trafficking regulator−2,758725,90E − 05
205282_atLRP8LDL receptor related protein 8−2,75490,000311
214440_atNAT1N-Acetyltransferase 1 (arylamine N-acetyltransferase)−2,685150,001777
219627_atZNF767PZinc finger family member 767, pseudogene−2,679570,00024
218984_atPUS7Pseudouridylate synthase 7 (putative)−2,675860,001308
206554_x_atSETMARSET domain and mariner transposase fusion gene−2,635360,002432
219779_atZFHX4Zinc finger homeobox 4−2,626240,001411
213103_atSTARD13StAR related lipid transfer domain containing 13−2,572190,002525
210138_atRGS20Regulator of G-protein signaling 20−2,559740,000415
204291_atZNF518AZinc finger protein 518A−2,543839,70E − 05
204651_atNRF1Nuclear respiratory factor 1−2,491470,003659
205408_atMLLT10Myeloid/lymphoid or mixed-lineage leukemia; translocated to, 10−2,489755,10E − 05
219581_atTSEN2tRNA splicing endonuclease subunit 2−2,453770,001774
218242_s_atSUV420H1Lysine methyltransferase 5B−2,446980,000754
203242_s_atPDLIM5PDZ and LIM domain 5−2,438510,001699
203868_s_atVCAM1Vascular cell adhesion molecule 1−2,435130,000761
220206_atZMYM1Zinc finger MYM-type containing 1−2,363620,008439
207616_s_atTANKTRAF family member associated NFKB activator−2,345670,000424
218303_x_atKRCC1Lysine-rich coiled-coil 1−2,345670,003187
218490_s_atZNF302Zinc finger protein 302−2,327850,001816
206876_atSIM1Single-minded family bHLH transcription factor 1−2,326240,001681
219128_atC2orf42Chromosome 2 open reading frame 42−2,286280,002926
212861_atMFSD5Major facilitator superfamily domain containing 5−2,270480,000823
218653_atSLC25A15Solute carrier family 25 member 15−2,256360,000562
206943_atTGFBR1Transforming growth factor beta receptor I−2,248560,025349
201995_atEXT1Exostosin glycosyltransferase 1−2,2470,000421
221430_s_atRNF146Ring finger protein 146−2,234570,001084
212286_atANKRD12Ankyrin repeat domain 12−2,22530,00029
219544_atBORABora, aurora kinase A activator−2,219140,000333
210455_atR3HCC1LR3H domain and coiled-coil containing 1 like−2,21760,0039
219459_atPOLR3BPolymerase (RNA) III subunit B−2,21760,000832
219078_atGPATCH2G-patch domain containing 2−2,199230,000723
204547_atRAB40BRAB40B, member RAS oncogene family−2,176480,001741
218791_s_atKATNBL1Katanin regulatory subunit B1 like 1−2,173470,001187
205173_x_atCD58CD58 molecule−2,171960,00022
204352_atTRAF5TNF receptor associated factor 5−2,168950,002659
204236_atFLI1Fli-1 proto-oncogene, ETS transcription factor−2,153970,005141
203072_atMYO1EMyosin IE−2,152480,000154
219904_atZSCAN5AZinc finger and SCAN domain containing 5A−2,148010,00144
219133_atOXSM3-Oxoacyl-ACP synthase, mitochondrial−2,122850,002424
205798_atIL7RInterleukin 7 receptor−2,112570,00506
205476_atCCL20C-C motif chemokine ligand 204,6139429,50E − 05
213497_atABTB2Ankyrin repeat and BTB domain containing 24,6235471,40E − 05
219179_atDACT1Dishevelled-binding antagonist of beta-catenin 14,6428169,00E − 06
219228_atZNF331Zinc finger protein 3314,7239716,00E − 06
213139_atSNAI2Snail family zinc finger 24,766731,40E − 05
218177_atCHMP1BCharged multivesicular body protein 1B4,8065441,00E − 05
203304_atBAMBIBMP and activin membrane-bound inhibitor4,8265763,00E − 06
201631_s_atIER3Immediate early response 34,8332713,00E − 06
218559_s_atMAFBv-maf avian musculoaponeurotic fibrosarcoma oncogene homolog B4,8702640,000468
220266_s_atKLF4Kruppel-like factor 4 (gut)4,8905610,00022
209211_atKLF5Kruppel-like factor 5 (intestinal)4,9245780,002036
209681_atSLC19A2Solute carrier family 19 member 24,9279925,90E − 05
205266_atLIFLeukemia inhibitory factor4,9553952,20E − 05
204790_atSMAD7SMAD family member 75,0735660,000283
221667_s_atHSPB8Heat shock protein family B (small) member 85,4226572,90E − 05
212665_atTIPARPTCDD-inducible poly(ADP-ribose) polymerase5,5250981,00E − 05
202935_s_atSOX9SRY-box 95,9711143,30E − 05
202023_atEFNA1Ephrin-A16,1645693,30E − 05
202393_s_atKLF10Kruppel-like factor 106,1945520,000195
213146_atKDM6BLysine demethylase 6B6,2031461,90E − 05
205193_atMAFFv-maf avian musculoaponeurotic fibrosarcoma oncogene homolog F6,29412,00E − 06
209457_atDUSP5Dual specificity phosphatase 56,6391571,30E − 05
206029_atANKRD1Ankyrin repeat domain 16,657590,008591
209283_atCRYABCrystallin alpha B6,7038970,000118
201693_s_atEGR1Early growth response 17,0567314,10E − 05
212099_atRHOBras homolog family member B7,3005240,000406
219682_s_atTBX3T-box 37,7221365,80E − 05
201473_atJUNBjun B proto-oncogene8,3224027,00E − 06
200664_s_atDNAJB1DnaJ heat shock protein family (Hsp40) member B18,5860822,00E − 05
205828_atMMP3Matrix metallopeptidase 38,7119761,90E − 05
201169_s_atBHLHE40Basic helix-loop-helix family member e408,8704050,00011
203665_atHMOX1Heme oxygenase 19,324330,000544
202643_s_atTNFAIP3TNF alpha induced protein 39,5731922,50E − 05
205207_atIL6Interleukin 610,182363,00E − 06
202388_atRGS2Regulator of G-protein signaling 210,253181,40E − 05
204472_atGEMGTP binding protein overexpressed in skeletal muscle10,80031,00E − 06
202149_atNEDD9Neural precursor cell expressed, developmentally down-regulated 911,065532,50E − 05
219480_atSNAI1Snail family zinc finger 111,704572,00E − 06
218839_atHEY1hes related family bHLH transcription factor with YRPW motif 112,075416,00E − 06
206115_atEGR3Early growth response 314,191941,20E − 05
204470_atCXCL1C-X-C motif chemokine ligand 117,618272,00E − 06
204621_s_atNR4A2Nuclear receptor subfamily 4 group A member 218,778370
209774_x_atCXCL2C-X-C motif chemokine ligand 219,027319,00E − 06
202859_x_atCXCL8C-X-C motif chemokine ligand 819,890391,00E − 06
202340_x_atNR4A1Nuclear receptor subfamily 4 group A member 120,749431,00E − 06
209189_atFOSFBJ murine osteosarcoma viral oncogene homolog23,360511,00E − 06
202672_s_atATF3Activating transcription factor 324,184320
202768_atFOSBFBJ murine osteosarcoma viral oncogene homolog B32,922450
207978_s_atNR4A3Nuclear receptor subfamily 4 group A member 343,804281,00E − 06
117_atHSPA6Heat shock protein family A (Hsp70) member 690,823890

3.3. Functional Analysis

To obtain a biological interpretation of the transcriptional signature of RSV-infected BEAS-2B cells and compare with those reported by previous studies, enrichment analysis was performed with the online platform for functional analysis InnateDB [21]. Based on a FDR < 0.1, DEGs identified for GSE3397 were enriched in pathways related to Chromatin organization, histone acetylation, signaling by NOTCH, IL1, Integrin-linked kinase signaling, EPO signaling pathway, VEGF signaling pathway, platelet degranulation, p73 transcription factor network, IL-7 signaling, p53 signaling pathway, and others (Figure 3(a) and Supplementary Data 1). Of interest, Huang and collaborators (2008) [13] reported gene overrepresentation within p53 signaling pathway, but only after 24 h following RSV infection of BEAS-2B cells. After 4 h following RSV infection, Huang and collaborators (2008) [13] only found a significant association with neuroactive ligand-receptor interaction pathway, which was not overrepresented in the present analysis. In contrast, DEGs resultant from dataset GSE6802 were enriched in pathways related to AP-1 transcription factor, ATF-2 transcription factor, IL-6 signaling, SMAD function, signaling by TGFBR, HIF-1α transcription factor, signaling by CD40/CD40L, signaling by MAPK, signaling by innate immune receptors, and others (Figure 3(b) and Supplementary Data 1). Some of those pathways as CD40 signaling are indeed commonly induced by a variety of viral respiratory infections [22], whereas several of those pathways could indicate novel directions for studying the host response against RSV. Six pathways were enriched by DEGs from both datasets, the EPO signaling pathway, FBXW7 Mutants and NOTCH1 in Cancer, IL1, p53 signaling pathway, p73 transcription factor network, and signaling by NOTCH1. The erythropoietin (EPO) gene is a primary target of HIF-1α transcription factor, whereas binding of HIF-1α to the EPO enhancer promoter region induces transcriptional programs that influence inflammation and infection processes [23]. In addition, expression of Dll4, a major NOTCH ligand, is upregulated in dendritic cells infected with RSV, whereas blockage of Dll4 in vivo increased hyperreactivity of airways and mucus secretion that impacted the pathology of the disease, showing a key role of signaling by NOTCH in the regulation of immunity against RSV [24]. Moreover, besides modulations of the p53 signaling pathway by infection of RSV in vitro [10, 13], this pathway was found to be upregulated in whole blood of children with lower respiratory tract infection by RSV [25]. Taken together, those data point to key pathways which can impact infections of human bronchial epithelial cells with RSV.

3.4. Meta-Analysis Based Biomarker Signature of RSV-Infected BEAS-2B Cells

To determine a unique transcriptional signature of BEAS-2B cells induced by early infection with RSV, common DEGs for both datasets were further identified. The analysis retrieved a list of seventeen common genes: ABCC4, ARMC8, BCLAF1, EZH1, FAM118A, FAM208B, FUS, HSPH1, KAZN, MAP3K2, N6AMT1, PRMT2, S100PBP, SERPINA1, TLK2, ZNF322, and ZNF337 (Figure 4). Despite particular features in expression data from both datasets, unsupervised hierarchical clustering analysis based on this signature revealed the formation of robust clusters between RSV-infected or uninfected BEAS-2B cells (Figure 4). Of note, human airway epithelial cells were shown to express ABCC4/MRP4, a transporter for uric acid and cAMP [26]. Mucosal production of uric acid was recently linked to particulate matter-induced allergic sensitization [26]; therefore RSV infection could trigger such a response and contribute to the development and severity of allergic responses to particulate matter [27]. Moreover, both ABCC4 and SERPINA1 are annotated into the platelet degranulation pathway (Figure 3(a)), suggesting a role in antiviral mechanisms from bronchial epithelial cells. After an initial encounter with RSV, the transcriptional activity of human bronchial epithelial cells is reprogrammed to counteract viruses and other pathogens [10], whereas MAP3K2 and ZNF322 are clearly involved on the activation and regulation of MAP kinase signaling pathway [28, 29]. Indeed, RSV infection leads to the activation of p38 MAPK [30] and c-JUN kinase pathway, which negatively regulates the production of TNF-α in human epithelial cells [31] and might contribute to virus evasion from an early immune response. Interestingly, the biosignature also included BCLAF1, a molecule involved in processes as apoptosis, transcription and processing of RNA, and export of mRNA from the nucleus [32]. However, this nuclear protein was also implicated as a viral restriction factor targeted to degradation by human cytomegalovirus [32]. Moreover, EZH1 was shown to be involved in the methylation of histone 3 at lysine 27 (H3K27) of the HIV provirus in resting cells [33] and could thus exert a significant function in infections with RSV, whereby other genes such as N6AMT1, FUS, and PRMT2 are also involved in protein methylation. Indeed, using coimmunoprecipitation and mass spectrometry, recent work demonstrated that RSV nucleoprotein (N) interacts with protein arginine N-methyltransferase 5 (PRMT5) [34], suggesting that PRMT2 could also interact with RSV proteins and play an important role during infections of human bronchial epithelial cells. Several of the genes identified in this study have been poorly studied in the context of RSV infection, whereby none of them was previously reported as a biomarker of infections by this virus. Of note, except for FAM208B and KAZN, analysis conducted by Smith and collaborators (2012) [22] which included both datasets (GSE3397 and GSE6802) also identified the significant modulation of the genes included in the biomarker signature identified herein.

4. Conclusions

The combined analysis of distinct datasets from BEAS-2B cells infected with RSV retrieved intriguing results, whereby using powerful statistical methods and assumptions this study identified a new set of biomarkers of early infection with RSV composed by seventeen genes: ABCC4, ARMC8, BCLAF1, EZH1, FAM118A, FAM208B, FUS, HSPH1, KAZN, MAP3K2, N6AMT1, PRMT2, S100PBP, SERPINA1, TLK2, ZNF322, and ZNF337. This transcriptional signature could be useful for the development of molecular diagnosis tools as well as future investigations of processes involved in host-pathogen interactions.

Competing Interests

The author declares that there are no competing interests regarding the publication of this paper.


The author is grateful to Dr. Fátima Pereira de Souza for critical comments on the paper. Luiz Gustavo Gardinassi was supported by scholarships from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP).

Supplementary Materials

Pathway enrichment analysis with the web-based platform InnateDB.

  1. Supplementary Material


  1. L. J. Anderson, R. A. Parker, and R. L. Strikas, “Association between respiratory syncytial virus outbreaks and lower respiratory tract deaths of infants and young children,” The Journal of Infectious Diseases, vol. 161, no. 4, pp. 640–646, 1990. View at: Publisher Site | Google Scholar
  2. A. R. Falsey and E. E. Walsh, “Respiratory syncytial virus infection in adults,” Clinical Microbiology Reviews, vol. 13, no. 3, pp. 371–384, 2000. View at: Publisher Site | Google Scholar
  3. V. F. Botosso, P. M. D. Zanotto, M. Ueda et al., “Positive selection results in frequent reversible amino acid replacements in the G protein gene of human respiratory syncytial virus,” PLoS Pathogens, vol. 5, no. 1, Article ID e1000254, 2009. View at: Publisher Site | Google Scholar
  4. L. G. A. Gardinassi, P. V. M. Simas, D. E. Gomes et al., “Diversity and adaptation of human respiratory syncytial virus genotypes circulating in two distinct communities: public hospital and day care center,” Viruses, vol. 4, no. 11, pp. 2432–2447, 2012. View at: Publisher Site | Google Scholar
  5. L. G. Gardinassi, P. V. Marques Simas, J. B. Salomão et al., “Seasonality of viral respiratory infections in southeast of Brazil: the influence of temperature and air humidity,” Brazilian Journal of Microbiology, vol. 43, no. 1, pp. 98–108, 2012. View at: Publisher Site | Google Scholar
  6. R. A. Tripp, “Respiratory syncytial virus (RSV) modulation at the virus-host interface affects immune outcome and disease pathogenesis,” Immune Network, vol. 13, no. 5, pp. 163–167, 2013. View at: Publisher Site | Google Scholar
  7. T. H. Mogensen, “Pathogen recognition and inflammatory signaling in innate immune defenses,” Clinical Microbiology Reviews, vol. 22, no. 2, pp. 240–273, 2009. View at: Publisher Site | Google Scholar
  8. G. Diamond, D. Legarda, and L. K. Ryan, “The innate immune response of the respiratory epithelium,” Immunological Reviews, vol. 173, pp. 27–38, 2000. View at: Publisher Site | Google Scholar
  9. R. J. Boyton and P. J. Openshaw, “Pulmonary defences to acute respiratory infection,” British Medical Bulletin, vol. 61, pp. 1–12, 2002. View at: Publisher Site | Google Scholar
  10. A. K. Mayer, M. Muehmer, J. Mages et al., “Differential recognition of TLR-dependent microbial ligands in human bronchial epithelial cells,” Journal of Immunology, vol. 178, no. 5, pp. 3134–3142, 2007. View at: Publisher Site | Google Scholar
  11. I. Martínez, L. Lombardía, B. García-Barreno, O. Domínguez, and J. A. Melero, “Distinct gene subsets are induced at different time points after human respiratory syncytial virus infection of A549 cells,” Journal of General Virology, vol. 88, no. 2, pp. 570–581, 2007. View at: Publisher Site | Google Scholar
  12. Y. Zhang, B. A. Luxon, A. Casola, R. P. Garofalo, M. Jamaluddin, and A. R. Brasier, “Expression of respiratory syncytial virus-induced chemokine gene networks in lower airway epithelial cells revealed by cDNA microarrays,” Journal of Virology, vol. 75, no. 19, pp. 9044–9058, 2001. View at: Publisher Site | Google Scholar
  13. Y.-C. T. Huang, Z. Li, X. Hyseni et al., “Identification of gene biomarkers for respiratory syncytial virus infection in a bronchial epithelial cell line,” Genomic Medicine, vol. 2, no. 3-4, pp. 113–125, 2008. View at: Publisher Site | Google Scholar
  14. R. Ihaka and R. Gentleman, “R: a language for data analysis and graphics,” Journal of Computational and Graphical Statistics, vol. 5, no. 3, pp. 299–314, 1996. View at: Google Scholar
  15. R. C. Gentleman, V. J. Carey, D. M. Bates et al., “Bioconductor: open software development for computational biology and bioinformatics,” Genome Biology, vol. 5, no. 10, article R80, 2004. View at: Publisher Site | Google Scholar
  16. L. Gautier, L. Cope, B. M. Bolstad, and R. A. Irizarry, “Affy—analysis of Affymetrix GeneChip data at the probe level,” Bioinformatics, vol. 20, no. 3, pp. 307–315, 2004. View at: Publisher Site | Google Scholar
  17. W. E. Johnson, C. Li, and A. Rabinovic, “Adjusting batch effects in microarray expression data using empirical Bayes methods,” Biostatistics, vol. 8, no. 1, pp. 118–127, 2007. View at: Publisher Site | Google Scholar | Zentralblatt MATH
  18. J. T. Leek, W. E. Johnson, H. S. Parker, A. E. Jaffe, and J. D. Storey, “The SVA package for removing batch effects and other unwanted variation in high-throughput experiments,” Bioinformatics, vol. 28, no. 6, Article ID bts034, pp. 882–883, 2012. View at: Publisher Site | Google Scholar
  19. M. E. Ritchie, B. Phipson, D. Wu et al., “Limma powers differential expression analyses for RNA-sequencing and microarray studies,” Nucleic Acids Research, vol. 43, article e47, 2015. View at: Publisher Site | Google Scholar
  20. G. R. Warnes, B. Bolker, L. Bonebakker et al., “gplots: various R programming tools for plotting data,” R Package Version, vol. 2, no. 4, 2009. View at: Google Scholar
  21. K. Breuer, A. K. Foroushani, M. R. Laird et al., “InnateDB: Systems biology of innate immunity and beyond—recent updates and continuing curation,” Nucleic Acids Research, vol. 41, no. 1, pp. D1228–D1233, 2013. View at: Publisher Site | Google Scholar
  22. S. B. Smith, W. Dampier, A. Tozeren, J. R. Brown, and M. Magid-Slav, “Identification of common biological pathways and drug targets across multiple respiratory viruses based on human host gene expression analysis,” PLoS ONE, vol. 7, no. 3, Article ID e33174, 2012. View at: Publisher Site | Google Scholar
  23. M. V. Sitkovsky, D. Lukashev, S. Apasov et al., “Physiological control of immune response and inflammatory tissue damage by hypoxia-inducible factors and adenosine A2A receptors,” Annual Review of Immunology, vol. 22, pp. 657–682, 2004. View at: Publisher Site | Google Scholar
  24. M. A. Schaller, R. Neupane, B. D. Rudd et al., “Notch ligand Delta-like 4 regulates disease pathogenesis during respiratory viral infections by modulating Th2 cytokines,” The Journal of Experimental Medicine, vol. 204, no. 12, pp. 2925–2934, 2007. View at: Publisher Site | Google Scholar
  25. A. Mejias, B. Dimo, N. M. Suarez et al., “Whole blood gene expression profiles to assess pathogenesis and disease severity in infants with respiratory syncytial virus infection,” PLoS Medicine, vol. 10, no. 11, Article ID e1001549, 2013. View at: Publisher Site | Google Scholar
  26. M. J. Gold, P. R. Hiebert, H. Y. Park et al., “Mucosal production of uric acid by airway epithelial cells contributes to particulate matter-induced allergic sensitization,” Mucosal Immunology, vol. 9, pp. 809–820, 2016. View at: Publisher Site | Google Scholar
  27. M. Barends, A. Boelen, L. de Rond et al., “Influence of respiratory syncytial virus infection on cytokine and inflammatory responses in allergic mice,” Clinical and Experimental Allergy, vol. 32, no. 3, pp. 463–471, 2002. View at: Publisher Site | Google Scholar
  28. Y. Li, Y. Wang, C. Zhang et al., “ZNF322, a novel human C2H2 Krüppel-like zinc-finger protein, regulates transcriptional activation in MAPK signaling pathways,” Biochemical and Biophysical Research Communications, vol. 325, no. 4, pp. 1383–1392, 2004. View at: Publisher Site | Google Scholar
  29. P. K. Mazur, N. Reynoird, P. Khatri et al., “SMYD3 links lysine methylation of MAP3K2 to Ras-driven cancer,” Nature, vol. 510, no. 7504, pp. 283–287, 2014. View at: Publisher Site | Google Scholar
  30. T. R. Meusel and F. Imani, “Viral induction of inflammatory cytokines in human epithelial cells follows a p38 mitogen-activated protein kinase-dependent but NF-κB-independent pathway,” The Journal of Immunology, vol. 171, no. 7, pp. 3768–3774, 2003. View at: Publisher Site | Google Scholar
  31. M. J. Stewart, S. B. Kulkarni, T. R. Meusel, and F. Imani, “c-Jun N-terminal kinase negatively regulates dsRNA and RSV induction of tumor necrosis factor-α transcription in human epithelial cells,” Journal of Interferon and Cytokine Research, vol. 26, no. 8, pp. 521–533, 2006. View at: Publisher Site | Google Scholar
  32. S. H. Lee, R. F. Kalejta, J. Kerry et al., “BclAF1 restriction factor is neutralized by proteasomal degradation and microRNA repression during human cytomegalovirus infection,” Proceedings of the National Academy of Sciences of the United States of America, vol. 109, no. 24, pp. 9575–9580, 2012. View at: Publisher Site | Google Scholar
  33. M. K. Tripathy, M. E. M. McManamy, B. D. Burch, N. M. Archin, and D. M. Margolis, “H3K27 demethylation at the proviral promoter sensitizes latent HIV to the effects of vorinostat in ex vivo cultures of resting CD4+ T cells,” Journal of Virology, vol. 89, no. 16, pp. 8392–8405, 2015. View at: Publisher Site | Google Scholar
  34. A. P. Oliveira, F. M. Simabuco, R. E. Tamura et al., “Human respiratory syncytial virus N, P and M protein interactions in HEK-293T cells,” Virus Research, vol. 177, no. 1, pp. 108–112, 2013. View at: Publisher Site | Google Scholar

Copyright © 2016 Luiz Gustavo Gardinassi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1057 Views | 457 Downloads | 2 Citations
 PDF  Download Citation  Citation
 Download other formatsMore
 Order printed copiesOrder
 Sign up for content alertsSign up