Home

Temporal Dynamics of Host Molecular Responses

1. cellular response to oxidative stress The superoxide dismutase E PLoS Genetics www plosgenetics org SOD1 and serine threonine kinase 25 STK25 or SOKI are markedly activated in Asx subjects contrasting to the transient suppression pattern 45 60 hpi in Sx hosts Figure 5D Figure S10 As SOD1 and STK25 both have been linked to anti oxidant stress response and reduced concentration of ROS 37 39 their sustained up regulation in Asx hosts highlights a host response signature unique to the Asx phenotype This signature may relate to the concomitant suppression of NLRP3 and ILIB in Asx individuals Collectively our data reveal a phenotypically divergent expression of NLR family genes and inflammasome signaling which may be related to the host anti oxidant response Distinct temporal kinetics of JAK STAT pathway and SOCS family genes reveals a potential anti inflammatory and viral control mechanism in Asx hosts A hallmark of host recognition of viral RNA is the activation of Janus kinase signal transducer and activator of transcription JAK STAT pathway which is crucial for the antiviral function of interferons However such activation is tightly controlled to limit the possibility of over stimulating inflammatory cytokine receptor signals As an integral component of the JAK STAT pathway the family of suppressor of cytokine signaling SOCS proteins have recently been reported to negatively regulate the response of August 2
2. tion We further present a statistical risk stratification model for estimating current disease state with potential forward risk assess ment capability These results are concordant with findings reported by Zaas et al that was limited to peak symptom time analysis Results Outline of overall analysis strategy A cohort of 17 healthy human volunteers Table S1 received intranasal inoculation of influenza H3N2 Wisconsin and 9 of these subjects developed mild to severe symptoms based on standardized symptom scoring 18 Gene expression profiles were measured on whole peripheral blood drawn from all subjects at an interval of 8 hours post inoculation hpi through 108 hpi A total of 267 gene expression profiles were obtained for all subjects at 16 time points including baseline 24 hpi As outlined in Figure S16 our analysis of the data consists of two parallel E PLoS Genetics www plosgenetics org Temporal Host Molecular Responses to Influenza A components 1 clinically uninformed unsupervised factor analysis using Bayesian Linear Unmixing BLU 19 2 clinically informed supervised pathway analysis using EDGE 20 and self organizing maps SOM 21 that leverages clinical and temporal covariates for increased statistical power The former establishes the existence of an ab initio molecular signature that strongly correlates to symptomatic clinical disease The later further reveals important host factors that delineat
3. 974 976 Pombo CM Bonventre JV Molnar A Kyriakis J Force T 1996 Activation of a human Ste20 like kinase by oxidant stress defines a novel stress response pathway Embo J 15 4537 4546 Yasukawa H Sasaki A Yoshimura A 2000 Negative regulation of cytokine signaling pathways Annu Rev Immunol 18 143 164 Rothlin CV Ghosh S Zuniga EI Oldstone MB Lemke G 2007 TAM receptors are pleiotropic inhibitors of the innate immune response Cell 131 1124 1136 Pothlichet J Chignard M Si Tahar M 2008 Cutting edge innate immune response triggered by influenza A virus is negatively regulated by SOCS1 and SOCS3 through a RIG I IFNAR1 dependent pathway J Immunol 180 2034 2038 Machado FS Johndrow JE Esper L Dias A Bafica A et al 2006 Anti inflammatory actions of lipoxin A4 and aspirin triggered lipoxin are SOCS 2 dependent Nat Med 12 330 334 Seki Y Hayashi K Matsumoto A Seki N Tsukada J et al 2002 Expression of the suppressor of cytokine signaling 5 SOCS5 negatively regulates IL 4 dependent STAT6 activation and Th2 differentiation Proc Natl Acad Sci US A 99 13003 13008 Korman BD Kastner DL Gregersen PK Remmers EF 2008 STAT4 genetics mechanisms and implications for autoimmunity Curr Allergy Asthma Rep 8 398 403 Ma A Koka R Burkett P 2006 Diverse functions of IL 2 IL 15 and IL 7 in lymphoid homeostasis Annu Rev Immunol 24 657 679 Schluns KS Lefrancois L 2003 Cytokine control of memory
4. MM et al 2009 A physical and regulatory map of host influenza interactions reveals pathways in HINL infection Cell 139 1255 1267 Andrejeva J Childs KS Young DF Carlos TS Stock N et al 2004 The V proteins of paramyxoviruses bind the IFN inducible RNA helicase mda 5 and inhibit its activation of the IFN beta promoter Proc Natl Acad Sci USA 101 17264 17269 Kang DC Gopalkrishnan RV Wu Q Jankowsky E Pyle AM et al 2002 mda 5 An interferon inducible putative RNA helicase with double stranded RNA E PLoS Genetics www plosgenetics org 16 Temporal Host Molecular Responses to Influenza A Ramsberg of Duke University for their helpful comments on earlier versions of this paper Author Contributions Conceived and designed the experiments AKZ TV CWW GSG BN Performed the experiments AKZ JBV CWW NC Analyzed the data YH MTM AOH Contributed reagents materials analysis tools AR ND PJW LC AOH Wrote the paper YH AKZ MTM SK CWW GSG AOH 28 29 30 31 32 34 36 37 38 39 50 51 dependent ATPase activity and melanoma growth suppressive properties Proc Natl Acad Sci US A 99 637 642 Hemmi H Kaisho T Takeuchi O Sato S Sanjo H et al 2002 Small anti viral compounds activate immune cells via the TLR7 MyD88 dependent signaling pathway Nat Immunol 3 196 200 Akira S Uematsu S Takeuchi O 2006 Pathogen recognition and innate immunity Cell 124 783 801 Pul
5. expression we extracted sets of genes that are most associated with differences between pairs of classes Table S4 When the expression profiles of these genes are plotted as heatmaps Figure 1C the contrasts in gene expression are striking For example the type I interferon antiviral response related genes IFI44L IFI27 GBPI RTP4 and OASI are among the most associated with differenti ating acute infection class 4 from the other 3 classes As another example note the contrast between complement component 3a receptor C3AR1 between Classes 2 and 3 exhibiting a marked change after inoculation in symptomatic subjects These genes are well known for their critical function in host immunity 6 23 24 This demonstrates both the strength of the genomic signature of acute infection and the utility of BLU factor analysis for ab initio discovery of this signature Identification of eight distinct virus mediated gene expression dynamics When we add clinical and temporal information about the samples to the analysis we can identify clusters of genes whose temporal expression patterns differentiate immune response of clinically asymptomatic from clinically symptomatic subjects Using EDGE with false discovery rate FDR significance level q value lt 0 01 we selected 5 076 genes whose temporal expres sion profiles differed significantly between Asx and Sx phenotypes Heatmaps of these 5 076 EDGE genes are shown in Figure S18 Next these 5 076
6. genes testing association with a clinical outcome Bioinformatics 20 93 99 Subramanian A Tamayo P Mootha VK Mukherjee S Ebert BL et al 2005 Gene set enrichment analysis a knowledge based approach for interpreting genome wide expression profiles Proc Natl Acad Sci US A 102 15545 15550 B hlmann P Yu B 2003 Boosting with the L2 loss regression and classification Journal of the American Statistical Association 98 324 339 Efron B 1979 Bootstrap Methods Another Look at the Jackknife The Annals of Statistics 7 1 26 August 2011 Volume 7 Issue 8 e1002234
7. haemagglutination inhibition HAI assay are not different from those of other Asx individuals Table S2 Figure 14 Figure S15 With all presented evidence supporting the activation of a robust Asx immune response our findings point to an important host factor that may lead to such Asx subclinical infections Shutting down protein synthesis helps control infection by inducing apoptosis of infected cells 60 62 Consistent with this we observed marked downregulation of protein biosynthesis and apoptosis related genes in Sx hosts at mid to late stages Figures S4 S5 S6 A similar lowering expression of ribosomal proteins has been reported in measles infected dendritic cells 63 What is surprising is the sustained upregulation of as many as 35 ribosomal proteins in only Asx subjects Figure 6C Figure S4 The increased ribosomal gene expression has been associated with peripheral blood lymphocytes 49 and our data also showed significant increase of white blood cells in Asx subjects Figure 17 Lacking strong PRRs activation and hence an absence of adaptive immune response these Asx hosts appeared to be capable of mounting a more potent cell mediated innate immune response than the symptomatic subjects Uncontrolled factors As our study mainly focuses on gene expression in whole peripheral blood it is possible that the changes observed in gene expression levels are at least partially due to changes in cell population However thi
8. independent of disease phenotype was performed using Fisher exact test Ho is rejected at significance level of 0 01 PDF August 2011 Volume 7 Issue 8 e1002234 Text S1 Supplementary methods and supplementary discus sions DOC Acknowledgments We gratefully acknowledge Retroscreen for their role in collecting the data and Stephanie Dobos Daphne Jones Anthony Moody and Elizabeth References l 2 20 21 22 23 24 25 26 27 Cox NJ Subbarao K 1999 Influenza Lancet 354 1277 1282 Carrat F Vergu E Ferguson NM Lemaitre M Cauchemez S et al 2008 Time lines of infection and disease in human influenza a review of volunteer challenge studies Am J Epidemiol 167 775 785 de Jong MD Simmons CP Thanh TT Hien VM Smith GJ et al 2006 Fatal outcome of human influenza A H5N1 is associated with high viral load and hypercytokinemia Nat Med 12 1203 1207 Palese P 2004 Influenza old and new threats Nat Med 10 S82 87 Kawai T Akira S 2007 TLR signaling Semin Immunol 19 24 32 Stetson DB Medzhitov R 2006 Type I interferons in host defense Immunity 25 373 381 Kawai T Sato S Ishii KJ Coban C Hemmi H et al 2004 Interferon alpha induction through Toll like receptors involves a direct interaction of IRF7 with MyD88 and TRAF6 Nat Immunol 5 1061 1068 Honda K Yanai H Negishi H Asagiri M Sato M et al 2005 IRF 7 is the master regulator of type I interferon dep
9. of genes that were selected by LogitBoost algorithm for each class pair more than 100 50 of the 200 bootstrap samples The average expression of these genes are shown in Figure 1C Supporting Information Figure S1 Temporal expression of Toll like receptor 7 pathway member genes Accompanying Figure 2c temporal expression are shown for TLR7 pathways genes n 11 including STAT1 IRF7 MyD88 TLR7 TNF CD40 IRF5 CD86 TRAF6 TBK1 and IFNAR 1 The expression intensities are averaged over subjects in Asx and Sx phenotypes and plotted on a log base 2 scale PDF Figure S2 Temporal expression of NLR family genes 1 cluster 7 gene PYD and CARD domain containing PYCARD or ASC 2 cluster 3 gene receptor interacting serine threonine kinase 2 RIPK2 3 cluster 2 gene caspase 5 CASP5 The expression intensities are plotted on a log base 2 scale PDF Figure S3 Increased temporal expression of antiviral RNA dependent eIF 2 alpha protein kinase EIF2AK2 or PKR in cluster 3 The expression intensities are plotted on a log base 2 scale PDF Figure S4 Phenotypically contrasting expression dynamics ribosomal protein synthesis related genes n 35 in cluster 6 The expression intensities are averaged over subjects in Asx and Sx phenotypes and normalized to have zero mean and unit standard deviation PDF Figure S5 Symptomatic specific temporal downregulation of cluster 4 genes n 9 that regulate programmed cell death a
10. on day 0 and day 7 with pre inoculation nAb independent of disease severity Critically the nAb titre increased over time in both Asx and Sx individuals Figure 12 This indicates a boosting effect of immunity and suggests that even if viral replication was inhibited enough viruses were detected by the Asx host immune system to cause expansion of Ab producing cells Secondly there was no apparent dosage effect subjects who received relatively lower amount of inoculation do not necessarily become more ill than individuals who received higher dose of virus We found no statistically significant dependence between disease outcome and inoculation dosage Figure S13A Furthermore the amount of viral shedding from the site of infection did not appear to differ among groups who received varying inoculation doses Figure S13B Thirdly Asx subjects presented dramatic transcriptional responses towards inoculation When their expression profiles were studied alone more than 3 000 genes showed significant post infection expression changes These changes do not correlate with the amount of virus detected Two subjects 3 and 17 who never yielded detectable virus lt 1 25 TCIDs5 9 mL in their E PLoS Genetics www plosgenetics org 11 nasal wash appeared to have the most significant temporal suppression of gene NLRP3 Table S2 Figure S14 Figure S15 Additionally the responses of two seroconverted Asx subjects 2 and 3 according to
11. pandemic virus induced upregulation of inflamma some components in a macaque model while avian H5N1 virus Vietnam 1203 04 caused increasing expression of NLR family genes in mice 50 51 In both cases the early and sustained upregulation of inflammasome component genes was directly associated with lethal or detrimental host responses Abnormal expression of NOD2 has been implicated in inflammatory bowel disease 52 53 Conversely it was shown in a study on chronic arthritis that Nod2 gene deficiency resulted in reduced joint inflammation and increased protection against early cartilage damage in mice 54 Our results provide new evidence for a much broader role played by NLR family genes during influenza viral infection that is likely to be shared by multiple viral strains August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A Expression Value Expression Value Expression Value 12 12 36 60 84 108 12 36 60 84 108 12 12 36 60 84108 12 36 60 84 108 Time hpi Time hpi Figure 5 Divergent expression patterns of Nod NACHT LRR NLRs family of genes from cluster 2 and cluster 3 with contrasting expression of anti oxidant stress genes SOD1 and STK25 or SOK1 A SOM cluster 3 genes nucleotide binding oligomerization domain containing 1 NOD1 and caspase 1 CASP1 display strong temporal upregulation in symptomatic subjects B SOM cluster 2 genes NOD2 and NLRP3 exhibit downregulation in Asx h
12. shedding and serological testing data for all human volunteers n 17 challenged with Influenza H3N2 viruses A Measure of viral titre isolated from nasal wash over a total of 9 days B Serological data on pre screening 24 hpi and 28 days PDF Table S3 Significance of monotonic trend of gene expression in SOM clusters For the genes in each SOM cluster Figure 1 we implemented the Jonkheere Terpstra JT test Text Sl of significance on Asx and Sx subjects respectively to test for monotonic increase or decrease of gene expression over time Columns 2 and 3 show p values associated with the null hypothesis that genes in the cluster have no monotonic trend Red colored entries indicate clusters having highly significant monotonic expression profiles for a particular phenotype PDF Table S4 Discriminatory genes selected by each logistic boosting model Genes are listed in decreasing order based on their discriminatory power in each model PDF Table S5 Comparison of genes identified by Zaas et al with significant genes in the present manuscript PDF Table S6 The proportions of primary white blood cell WBC subtypes are similar between Asx and Sx White blood cells counts were obtained daily through standard laboratory workout Phenotype specific average percentage of cell subpopulation were computed using Tukey s biweight robust M estimator The null hypothesis Hy the frequency distribution of WBC subtypes is
13. which point every one of the 35 genes becomes highly expressed Individually all genes showed increased expression trend Figure S4 This trend can be seen at as early as 5 hpi and as late as 108 hpi In contrast Sx subjects showed sustained down regulation of the same set of genes with lowest expression level at 60 hpi This decreasing trend continues until 84 hpi which coincides with the peak symptom time observed in symptomatic subjects Figure 3 In addition these 35 E PLoS Genetics www plosgenetics org genes include 53 of 13 genes whose expression are characteristic of peripheral blood lymphocytes Figure 6D 49 suggesting prominent presence of lymphocytes in the blood of Asx subjects during infection This is further supported by the increased number of whole blood leukocytes in Asx subjects Figure S17 Given the markedly contrasting trends observed between Asx and Sx phenotypes we conclude that Asx hosts responded differently to the viral insult by inducing leukocyte response with enhanced cellular protein biosynthesis Discussion Pathogenic influenza A viral infection is a complex and dynamic process that involves various components of the host immune system at different stages of infection in response to virus induced physiological changes Dissecting the temporal host response to invading viruses and subsequent symptomatic disease process are crucial for studying disease pathogenesis and related host factors Equal
14. 011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A A _Toll like receptor signaling pathway 12 11 10 pN 9 8 e 7 gt 6 e S 4 l 3 2 IFNAR1 1 0 E Pn ANASA NNAS AAAA A 12 0 5 12 21 29 36 45 53 60 69 77 84 93 101 108 hpi B Ou 5 S H H 2 1 ees S 8 5 Gs i ci deat S 8 0 4 A Oe A ae e i i ch v i E Bei die i ily dt x u es C 2 ZS E Ki YN N Nd oe x lt l D eb 2 Ss E Ki Y YN A k Hii diam e PA SS g FRA Sg 5 g AR Wi nz 42 12 36 60 84 108 12 36 60 84 108 Time hpi Time hpi E PLoS Genetics www plosgenetics org 8 August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A Figure 4 Similar expression dynamics of TLR7 pathway effector genes in cluster 3 A Significance of association p value between Toll like receptor TLR pathway and overall symptom severity Significant positive association between TLR pathway genes and symptom severity is shown at 53 hpi top left The temporal expression of representative significant genes on TLR pathway that are related to pattern recognition TLR7 and RIG I B antiviral myxovirus resistant 1 MX1 and 2 5 oligoadenylate synthetase 1 OAS1 C and pro inflammatory TNF and SIGLEC 1 D The expression intensities are plotted on a log base 2 scale and all genes are differentially expressed between Asx and Sx q value lt 0 0001 EDGE doi 10 1371 journal
15. 228 oe 1 br P 0 gt 2 0 Figure 2 Self organizing map clusters show distinct transcriptional dynamics in Influenza H3N2 Wisconsin virus challenge study A Polar plots of the 8 SOM clusters and their associated gene expression patterns Each segment plot represents the prototype of a cluster Individual time points are scaled and ordered in sequence and phenotype around the circle Specifically the temporal expression of Asx resides on the top portion of the circle while Sx expression occupies the bottom half Each phenotype s expression values are placed in time sequence with E PLoS Genetics www plosgenetics org 5 August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A time increasing in the counterclock wise direction inside its own half circle The degrees of angle are equally divided among segments within the circular plot The different lengths of radii of the segments represent the deviation of a time point from the average expression level of the complete time course B Heatmaps of EDGE estimated temporal profiles of the top 5 genes from each SOM cluster EDGE averages over Asx left and Sx subjects right and smoothes over time using a fitted cubic spline Genes in each cluster are ordered in decreasing order of EDGE significance level See Figure 18 for heatmaps of all significant genes found by EDGE C Centroids of each SOM cluster show individual cluster average expression profile
16. AP RTP4 GNG7 OAS1 C13orf18 IFI27 BLVRA ADAMTS5 IFI44L SMAD1 RTP4 3vs4 EGR2 ADRBK2 FLJ12529 2vs3 2vs4 Figure 1 The BLU genetic signature correlates strongly with disease severity and yields early and late stage risk stratification model A The scores of the top ranked factor detected by the unsupervised BLU factor analysis method Each microarray sample is represented by one square cell of the image and ordered by phenotype and subject row wise and increasing time column wise Color palette is coded according E PLoS Genetics www plosgenetics org 3 August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A to the enrichment factor score determined by BLU The higher the score the warmer red color representation of the sample The numbers 1 to 4 are the disease state class designation determined by BLU 3 and 4 and inoculation time 1 and 2 The boundary between 3 and 4 occurs at samples that are labeled To denoting the critical transition point onset time of Sx subject transcriptome profiles The grey color absolute O loading corresponds to samples that were not assayed B Clinical symptom chart of corresponding subjects rows and times columns that are ordered in the same manner as A C Heatmap of groups of genes that are most highly associated with differences between pairs of classes according to a logistic regression model For purposes of visualization the heatmaps s
17. IL1ORB TNFRSF14 TNFSF10 TNFSF12 BTK RNASE2 C3AR1 CYBB FASLG APOL3 ANXA2 1F135 IFIT1 IFIT3 IFITM1 IFITM3 4 1175 oxidative stress ca induced T cell CCL5 RPS6KA5 ACTG1 CUL3 PRKC GENES C JUN PIK3 family MAP2K4 apoptosis iCOS signaling CD3E CD247 CD40LG CAMK4M IL2RB ITK ITPR1 ITPR3 LAT NFATC1 NFATC3 ICOS FYN 5 228 antigen presentation innate immune CD97 THBD DDX17 IL1R2 ORM1 TREM1 AOC3 FOXO3 IL1R1 IL1RAP response AQP9 CA4 CAMK1D 6 1326 protein synthesis oxidative stress RNA SOCS2 SOCS5 SOD1 SOK1 RPL3 EIF3 FAMILY GENES CCR7 RPS9 trafficking JAK STAT signaling RPS14 RPL22 C1QBP DDX21 DDX50 ICOS 7 228 natural killer cell signaling cell apoptosis SIGLEC7 ASC SHC1 MAPK7 KIR2DL1 KIR2DS4 KIR3DL1 SERPINF1 RAC1 CD4 CX3CR1 HLA G TNFRSF1B ITGB2 CTSD 8 171 cell morphology cell signaling EIF2AK1 LY96 BCL2L1 KRAS PIM1 TGM2 RGS1 PKN2 doi 10 1371 journal pgen 1002234 t001 E PLoS Genetics www plosgenetics org August 2011 Volume 7 Issue 8 e1002234 Disease Symptoms Flu001 FIUDOS FOE Fiuoo7 FOE Flu010 FOTZ FOTZ Fiu015 Jackson Score o 25 50 75 100 125 150 Time hpi 8 9 Symptomatic Sx Flu001 FIu005 Flu006 Flu007 Flu008 Flu010 D Flu012 50 Flu013 Flu015 Expression Value 40 O 5 12 22 29 36 46 53 60 70 77 84 94 101108 89 Asymptomatic Asx FL
18. OPEN Q ACCESS Freely available online PLOS cenetics Temporal Dynamics of Host Molecular Responses Differentiate Symptomatic and Asymptomatic Influenza A Infection Yongsheng Huang Aimee K Zaas Arvind Rao Nicolas Dobigeon Peter J Woolf Timothy Veldman N Christine ien Micah T McClain Jay B Varkey Bradley Nicholson Lawrence Carin Stephen Kingsmore Christopher W Woods Geoffrey S Ginsburg Alfred O Hero Wille 1 Center for Computational Biology and Bioinformatics University of Michigan Ann Arbor Michigan United States of America 2 Department of Statistics University of Michigan Ann Arbor Michigan United States of America 3 Institute for Genome Sciences and Policy Duke University Durham North Carolina United States of America 4 Department of Medicine Duke University Durham North Carolina United States of America 5 Lane Center for Computational Biology Carnegie Mellon University Pittsburgh Pennsylvania United States of America 6 IRIT INP ENSEEIHT University of Toulouse Toulouse France 7 Department of Biomedical Engineering University of Michigan Ann Arbor Michigan United States of America 8 Department of Chemical Engineering University of Michigan Ann Arbor Michigan United States of America 9 Department of Medicine School of Medicine Emory University Atlanta Georgia United States of America 10 Department of Electrical and Computer Engineering Duke Univ
19. T cell development and survival Nat Rev Immunol 3 269 279 Sun JC Lehar SM Bevan MJ 2006 Augmented IL 7 signaling during viral infection drives greater expansion of effector T cells but does not enhance memory J Immunol 177 4458 4463 Palmer C Diehn M Alizadeh AA Brown PO 2006 Cell type specific gene expression profiles of leukocytes in human peripheral blood BMC Genomics 7 115 Cilloniz C Pantin Jackwood MJ Ni C Goodman AG Peng X et al 2010 Lethal dissemination of H5N1 influenza virus is associated with dysregulation of inflammation and lipoxin signaling in a mouse model of infection J Virol 84 7613 7624 Cilloniz C Shinya K Peng X Korth MJ Proll SC et al 2009 Lethal influenza virus infection in macaques is associated with early dysregulation of August 2011 Volume 7 Issue 8 e1002234 52 53 54 55 56 57 58 59 60 61 inflammatory related genes PLoS Pathog 5 el000604 doi 10 1371 journal ppat 1000604 Abraham C Cho JH 2009 Inflammatory Bowel Disease N Engl J Med 361 2066 2078 Hugot JP Chamaillard M Zouali H Lesage S Cezard JP et al 2001 Association of NOD2 leucine rich repeat variants with susceptibility to Crohn s disease Nature 411 599 603 Joosten LA Heinhuis B Abdollahi Roodsaz S Ferwerda G Lebourhis L et al 2008 Differential function of the NACHT LRR NLR members Nodl and Nod2 in arthritis Proc Natl Acad Sci U S A 105 9017 9022 Vand
20. UOO2 om FLU003 8 5 FLU004 20 FLU009 8 3 FLUO11 8 1 FLU014 10 FLU016 7 9 FLU017 8 7 Expression Value 7 7 DS 12 22 29 36 46 53 60 70 77 84 94 101108 Time hpi Earache Malaise 0 00 Cough Shortness of Breath 0 04 Headache 0 00 Muscles and or Joint Pain 0 00 Symptom Score Overall Temporal Host Molecular Responses to Influenza A Symptoms som som2 SOM3 SOM4 SOM5 SOM6 SOM7 SOM8 Runny Nose Stuffy Nose Sneezing Sore Throat 0 15 0 38 0 25 0 07 0 12 0 35 0 13 0 05 0 08 0 32 0 19 0 00 0 13 0 16 0 01 0 18 0 11 0 26 0 27 0 12 0 25 0 07 0 02 0 14 0 20 0 36 0 22 0 07 0 01 0 04 0 35 0 00 0 00 0 32 0 06 0 00 0 09 0 41 0 17 0 00 H3N2 Interacting Human Proteins D H1N1 Interacting Human Proteins SOM Clusters Clusters Figure 3 Strong correlation of molecular signature with disease severity A Clinical symptom scores of symptomatic subjects with individuals represented by curves in different colors B Cluster 3 gene expression of symptomatic subjects top and Asx subjects bottom C The correlation coefficients total variance explained between standardized symptom scores and SOM clusters D Cluster wise distribution of genes encoding proteins that are known to interact with H1N1 PR8 and H3N2 Udorn viruses The proportion of genes encoding virus human interaction proteins in a cluster is compared to the relative size of that cluster for determining sign
21. and corresponding two standard deviations The statistical significance of phenotype specific trend of expression monotonicity can be found in Table S3 hpi hours post inoculation doi 10 1371 journal pgen 1002234 g002 TLR signaling pathway including MyD88 TRAF6 and STAT 1 As a group they showed an aggregated effect that is significantly associated with the symptomatic disease This association reaches statistical significance p lt 0 05 Globaltest at 53 hpi with an increasing trend appearing as early as 36 hours before peak symptom time By 93 hpi the association attains its maximum level of significance with all 11 member genes significantly upre gulated Figure 4A Figure S1 The activation of PRRs by viral ligands triggers downstream signaling cascades that include both antiviral and inflammatory responses In line with this cluster 3 contains many such downstream effector genes that were fully activated with similar dynamics Several interferon stimulated antiviral genes such as MSL OAS1 RSAD2 PKR exhibit Sx specific significant tem poral activation beginning at 36 45 hpi Figure 4C Figure 83 Figure 9 This increase persists many hours beyond symptom peak time suggesting non rescinding efforts in viral resolution by the host It is noteworthy that none of the type I interferon genes themselves is differentially expressed between the Sx and Asx phenotypes Similarly cluster 3 also contains many elements of the inflam
22. but can still do activities and bothersome and cannot do daily activities For all cohorts modified Jackson scores were tabulated to determine if subjects became symptomatic from the respiratory viral challenge A modified Jackson score of gt 6 over the first five days period was the primary indicator of successful viral infection 18 64 and subjects with this score were denoted as Symptomatic Sx Viral titers from daily nasopharyngeal washes were used as corrobora tive evidence of successful infection using quantitative PCR Table S2 18 64 65 Subjects were classified as Asymptomatic if the Jackson score was less than 6 over the first five days of observation and viral shedding was not documented after the first 24 hours subsequent to inoculation Successful inoculation in Asx hosts was further validated by analysis of multimodal data including serum neutralizing antibody and haemagglutination inhibition titers For additional evidence see discussion in Text Sl Standardized symptom scores were tabulated at the end of each study to determine attack rate and time of maximal symptoms time T The clinical disease is mild only a single fever was observed Immune activation assays such as antibody response over the full E PLoS Genetics www plosgenetics org 13 time course of the challenge study were not available for our analysis Biological sample collections During the challenge study subjects had
23. contribute to the pathogenesis because their significant increase preceded disease manifestation by 36 hours In subclinical asymptomatic hosts we discovered strong transcriptional regulation of genes involved in inflammasome activation genes encoding virus interacting proteins and evidence of active anti oxidant and cell mediated innate immune response Taken together our findings offer insights into influenza virus induced patho genesis and provide a valuable tool for disease monitoring and management in natural environments The peripheral blood contains key elements of the immune system and the circulating immune cells recruited by the host in response to viral infection and virus induced tissue damage provides a global view of the host immune response Thus we hypothesized that it can be used to monitor the temporal dynamics of host virus interactions Analyzing whole genome gene expres sion profiles from healthy human subjects challenged with influenza H3N2 Wisconsin we studied the full temporal spectrum of virus mediated disease dynamics Going beyond the peak symptom time analysis reported in Zaas et al 17 this report offers an hour by hour detailed view of host immune response as a continuum spanning the time from exposure to peak symptom manifestation Utilizing biological and mathematical models we highlight key immune response events representing potential factors that determine the pathogenicity of influenza viral infec
24. ctious and can cause acute respiratory illness in human hosts Infected hosts present a variety of clinical symptoms including fever runny nose sore throat myalgias and malaise with potentially more serious complications such as viral pneumonia 1 Many hosts also withstand com parable level of viral msult with little or no overt symptoms exhibiting a higher degree of tolerance 2 3 Clearly these asymptomatic infected hosts are able to control and eradicate viral threats more effectively than those who become symptomatic Given the dynamic nature of viral infection it is now recognized that interactions between hosts and viruses play a crucial role in determining the presence and absence of symptoms 4 This leads to an interesting question _ what are the principal factors associated with such divergent disease outcome E PLoS Genetics www plosgenetics org In recent years seminal studies on the sensing of pathogens by pattern recognition receptors PRRs and their related signaling cascades have advanced our understanding of innate immunity 5 10 Many elegant experimental analyses have further eluci dated the mechanistic activation and modulation of host response to invading pathogens 11 16 By design however host re sponses in these experimental conditions are often characterized for individual cells via cell culture or they represent a snapshot of the immune response pertaining to a limited number of time points The comp
25. e 25 STK25 or SOK1 The expression intensities are plotted on a log base 2 scale PDF Figure S11 Temporal expression of genes from the family of suppressor of cytokine signaling SOCS including cluster 2 gene SOCS3 and cluster 6 gene SOCS5 The expression intensities are plotted on a log base 2 scale PDF Figure S12 Neutralizing antibody nAb measure prior to inoculation shows no significant phenotypic difference and is not correlated with disease outcome A B nAb of all subjects at Day 0 A and day 7 B No difference were observed between Asx and Sx on both days non parametric rank test C No evident correlation between nAb on Day 0 and maximum Jackson standardized score A linear regression fit of score on nAb readings is shown in dark black line Correlation test was performed using Spearman test D nAb increased in both Asx and Sx subjects from day 0 to day 28 x No sample available on day 28 PDF Figure S13 The infection outcome and viral load are indepen dent of the dosage of viral inoculation A There is no significant correlation between disease outcome and inoculation dosage p value 0 2299 Fisher s exact test Each bar represents a randomized group of four to five subjects receiving a varying dose of Influenza A virus inoculation on day 0 Supplementary Materials Within each group subjects are divided into clinically determined symptomatic red and asymptomatic blue subgroups B Viral sheddi
26. e important for the generation of Thl responses by repressing IL 4 induced signals that promote Th2 differentiation 44 In addition we observed a significant positive association of interleukin 7 IL7 and STAT4 Figure 6B Of these STAT4 transduces IL12 and IFN A cytokine signals in T cells and monocytes 45 whereas IL7 is critical for proper T cell response and expansion during viral infection 46 48 Taken together the distinct expression patterns of SOCS family genes and related JAK STAT signaling suggest possible early involve ment of Thl type adaptive immune response in Asx hosts with no sign of excessive inflammation Ribosomal protein synthesis genes are upregulated in Asx subjects as compared to Sx subjects In addition to expression changes in magnitude genes in clusters 2 Agr Za and 6 A Si also exhibit directional contrast between two phenotypes As the largest cluster with a total of 1 326 member genes cluster 6 contains genes with expression profiles similar to those of SOCS2 and SOCS5 Among them we found an unusual saturation of genes related to ribosomal protein synthesis Out of 47 significant genes in this pathway 35 76 of them are located in cluster 6 p value lt 0 0001 test Together these 35 genes correlated positively with Asx phenotype p value lt 0 05 Globaltest and their expression increased over the course of the study Figure 6C Such association emerges at 45 53 hpi and peaks at 60 hpi at
27. e time courses of designated symptomatic Sx and asymptomatic Asx subjects A genomic signature discriminates between early and late stages of disease Symptomatic infection exhibits a distinct time evolving mole cular signature This signature is sufficiently strong that a clinically uninformed factor analysis method is able to pick it up without using any clinical phenotype information such as disease outcome subject or time labels For this analysis we used the BLU factor analysis method described in the Methods section Figure 1A shows a heatmap of the linear combination BLU factor score of genes in this signature where for visualization we have arranged the samples in a matrix whose rows and columns are organized according to clinical phenotype of the subject and sample time The image of the BLU factor score shown in Figure 1A bears striking resemblance to the standardized clinical symptom obser vation matrix in Figure 1B The BLU factor score signature is sufficiently strong that application of a threshold to the post inoculation part of the heatmap in Figure 1A perfectly divides the subjects into asympto matic subjects Class 2 and symptomatic subjects before onset Class3 and after onset Class 4 of acute infection The selection of the threshold was based on the pre inoculation samples Class 1 and is described in the Methods section Then using logistic regression 22 as an association measure between class label and gene
28. emporal Host Molecular Responses to Influenza A Clemens MJ Elia A 1997 The double stranded RNA dependent protein kinase PKR structure and function J Interferon Cytokine Res 17 503 524 Zilliox MJ Parmigiani G Griffin DE 2006 Gene expression patterns in dendritic cells infected with measles virus compared with other pathogens Proc Natl Acad Sci US A 103 3363 3368 Turner RB 2001 Ineffectiveness of intranasal zinc gluconate for prevention of experimental rhinovirus colds Clin Infect Dis 33 1865 1870 Barrett B Brown R Voland R Maberry R Turner R 2006 Relations among questionnaire and laboratory measures of rhinovirus infection Eur Respir J 28 358 363 Bolstad BM Irizarry RA Astrand M Speed TP 2003 A comparison of normalization methods for high density oligonucleotide array data based on variance and bias Bioinformatics 19 185 193 Faraway J 2004 Linear Models with R Chapman and Hall H ssjer O 2008 On the coefficient of determination for mixed regression models Journal of Statistical Planning and Inference 138 3022 3038 Hero A Fleury G 2004 Pareto optimal methods for gene ranking Journal of VLSI Signal Processing 38 259 275 Benjamini Y Hochberg Y 1995 Controlling the false discovery rate a practical and powerful approach to multiple testing Journal of the Royal Statistical Society Series B 57 289 300 Goeman JJ van de Geer SA de Kort F van Houwelingen HC 2004 A global test for groups of
29. endent immune responses Nature 434 7712 777 Takaoka A Yanai H Kondo S Duncan G Negishi H et al 2005 Integral role of IRF 5 in the gene induction programme activated by Toll like receptors Nature 434 243 249 Yamamoto M Sato S Hemmi H Hoshino K Kaisho T et al 2003 Role of adaptor TRIF in the MyD88 independent toll like receptor signaling pathway Science 301 640 643 Ichinohe T Lee HK Ogura Y Flavell R Iwasaki A 2009 Inflammasome recognition of influenza virus is essential for adaptive immune responses J Exp Med 206 79 87 Yoneyama M Kikuchi M Natsukawa T Shinobu N Imaizumi T et al 2004 The RNA helicase RIG I has an essential function in double stranded RNA induced innate antiviral responses Nat Immunol 5 730 737 Zhu Q Egelston C Vivekanandhan A Uematsu S Akira S et al 2008 Toll like receptor ligands synergize through distinct dendritic cell pathways to induce T cell responses implications for vaccines Proc Natl Acad Sci U S A 105 16260 16265 Fenner JE Starr R Cornish AL Zhang JG Metcalf D et al 2006 Suppressor of cytokine signaling 1 regulates the immune response to infection by a unique inhibition of type I interferon activity Nat Immunol 7 33 39 Ryo A Tsurutani N Ohba K Kimura R Komano J et al 2008 SOCS1 is an inducible host factor during HIV 1 infection and regulates the intracellular trafficking and stability of HIV 1 Gag Proc Natl Acad Sci US A 105 294 299 Pro
30. ent All volunteers were without recent influenza like illness in the preceding 45 days tested influenza A H3N2 antibody negative by HAI at pre inoculation screening and had not been vaccinated with a seasonal influenza vaccine within the preceding 3 years On day of inoculation a dose of 10 TCIDso Influenza A manufactured and processed under current good manufacturing practices CGMP by Bayer Life Sciences Vienna Austria was inoculated intranasally per standard protocol at a varying dose 1 10 1 100 1 1000 1 10000 with four to five subjects receiving each dose Subjects were not released from quarantine until after the 216th hour Blood and nasal lavage collection continued throughout the duration of the quarantine All subjects received oral oseltamivir Roche Pharmaceuticals 75 mg by mouth twice daily prophylaxis at day 6 following inoculation All patients were tested negative by rapid antigen detection BinaxNow Rapid Influenza Antigen Inverness Medical Innovations Inc at time of discharge All exposures were ap proved by the relevant institutional review boards and conducted according to the Declaration of Helsinki Case definitions Symptoms were recorded twice daily using standardized symptom scoring 2 The modified Jackson Score requires subjects to rank symptoms of upper respiratory infection stuffy nose scratchy throat headache cough etc on a scale of 0 3 of no symptoms just noticeable bothersome
31. ermeer M Thomas A Kamimoto L Reingold A Gershman K et 2009 Role of statins in preventing death among patients hospitalized with lab confirmed influenza infections Infectious Diseases Society of America Annual Meeting 706 p Kofler S Nickel T Weis M 2005 Role of cytokines in cardiovascular diseases a focus on endothelial responses to inflammation Clin Sci Lond 108 205 213 Floyd RA Hensley K Jaffery F Maidt L Robinson K et al 1999 Increased oxidative stress brought on by pro inflammatory cytokines in neurodegenerative processes and the protective role of nitrone based free radical traps Life Sci 65 1893 1899 Schwarz KB 1996 Oxidative stress during viral infection a review Free Radic Biol Med 21 641 649 Allen IC Scull MA Moore CB Holl EK McElvania TeKippe E et al 2009 The NLRP3 inflammasome mediates in vivo innate immunity to influenza A virus through recognition of viral RNA Immunity 30 556 565 Castelli JC Hassel BA Wood KA Li XL Amemiya K et al 1997 A study of the interferon antiviral mechanism apoptosis activation by the 2 5A system J Exp Med 186 967 972 Samuel CE Kuhen KL George CX Ortega LG Rende Fournier R et al 1997 The PKR protein kinase an interferon inducible regulator of cell growth and differentiation Int J Hematol 65 227 237 E PLoS Genetics www plosgenetics org 17 62 63 64 65 66 67 68 69 70 71 72 73 74 T
32. ersity Durham North Carolina United States of America 11 Center for Pediatric Genomic Medicine Children s Mercy Hospital Kansas City Missouri United States of America 12 Department of Electrical Engineering and Computer Science University of Michigan Ann Arbor Michigan United States of America Abstract Exposure to influenza viruses is necessary but not sufficient for healthy human hosts to develop symptomatic illness The host response is an important determinant of disease progression In order to delineate host molecular responses that differentiate symptomatic and asymptomatic Influenza A infection we inoculated 17 healthy adults with live influenza H3N2 Wisconsin and examined changes in host peripheral blood gene expression at 16 timepoints over 132 hours Here we present distinct transcriptional dynamics of host responses unique to asymptomatic and symptomatic infections We show that symptomatic hosts invoke simultaneously multiple pattern recognition receptors mediated antiviral and inflammatory responses that may relate to virus induced oxidative stress In contrast asymptomatic subjects tightly regulate these responses and exhibit elevated expression of genes that function in antioxidant responses and cell mediated responses We reveal an ab initio molecular signature that strongly correlates to symptomatic clinical disease and biomarkers whose expression patterns best discriminate early from late phases of infection Our results
33. establish a temporal pattern of host molecular responses that differentiates symptomatic from asymptomatic infections and reveals an asymptomatic host unique non passive response signature suggesting novel putative molecular targets for both prognostic assessment and ameliorative therapeutic intervention in seasonal and pandemic influenza Citation Huang Y Zaas AK Rao A Dobigeon N Woolf PJ et al 2011 Temporal Dynamics of Host Molecular Responses Differentiate Symptomatic and Asymptomatic Influenza A Infection PLoS Genet 7 8 e1002234 doi 10 1371 journal pgen 1002234 Editor Nicholas J Schork University of California San Diego and The Scripps Research Institute United States of America Received January 5 2011 Accepted June 28 2011 Published August 25 2011 Copyright 2011 Huang et al This is an open access article distributed under the terms of the Creative Commons Attribution License which permits unrestricted use distribution and reproduction in any medium provided the original author and source are credited Funding This research was partially supported by a grant from the DARPA PHD Program The funders had no role in study design data collection and analysis decision to publish or preparation of the manuscript Competing Interests The authors have declared that no competing interests exist E mail geoffrey ginsburg duke edu GSG hero eecs umich edu AOH Introduction Influenza viruses are highly infe
34. gene expression profiles were grouped into clusters based on using SOM applied jointly to the Sx and Asx phenotypes A total of eight clusters were identified and their associated centroids are shown in Figure 2A and 2C as polar and linear plots of expression over time Heatmaps of gene expression August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A Flu002 Fiu0021 0121010101010101010101010 Flu003 Flu003 0 0 0 0 0 0 0 0 0 0 0 0 0 Flu004 Fluo04 o o o o o o o o o o o o o Asx Flu0o9 Fliu009lototo to to Flu011 Flu0111010101010 010101010 Flu014 Flu014101010101010101010 Flu016 Flu0i6 01 01 01 0 0 0 01010101010 Flu017 Flu017 0 0 0 0 0 0 0 0 0 0 0 0 Flu001 Flu001 0 o Flu005 Flu005 0 0 Flu006 Flu0061 01010 Flu007 Flu00710101010 Sx Fludos Flu008 0 0 0 0 0 Flu010 Flu010 0 0 0 0 0 0 Flu012 Fluo12 o o o o o o Flu013 Flu013 o to to to to to Flu015 Flu0151 010 DR 42 0 5 12 21 529 36 45 5 53 60 69 577 84 93 5101 108 hpi 1 121 0 116126 139 150 162174 186 1 98 1110 1221144 0 Asx 4080 Sx 108 hpi lt 1 0 gt 1 _ SS H Fluo17 O OOOO OOOO OOOO FLT3 EE SLC26A4 Bo FKBP5 LI ppm NEB SLC25A20 TLE1 KIFIB CIRBP 5 GBP1 JUS ae TLE1 SMAD1 EI IFI44L L ACP5 mi TUBB2A BEB E IP LPAR1 CL LOC93349 EU E 115648_at mn 1vs2 1vs3 1vs4 MYOM2 HLA DQA1 C3AR1 CTH CNKSRI LOC283345 TOMM34 UAPIL1 GSTT1 HRASLS3 GM2A _ WDR41 UTS2 PRSS21 IL18R
35. how gene expression profiles that are averaged over Asx left and Sx right phenotypes and are smoothed over time using the same cubic spline fitting method as used for heatmaps shown in Figure 2B doi 10 1371 journal pgen 1002234 g001 are shown for the top 5 genes in each SOM cluster Figure 2B These eight clusters decompose temporal host response into eight distinct classes of differential expression dynamics revealing divergent trends in asymptomatic and symptomatic responses over time The contrasts in expression patterns between pheno types are all statistically significant g value lt 0 01 Figure 20 Most clusters show significant monotonic increase or decrease in expression over time in Asx or Sx phenotypes Table 3 For Sx subjects we define three stages of infection early 0 12 hpi middle 12 43 hpi and late gt 45 hpi Collectively clusters 2 3 4 and 6 contain more than 78 of all significant genes and highlight the sharp contrasts in expression dynamics between phenotypes Although the discussion below focuses on these four clusters pathway enrichment analysis indi cates that genes from all eight clusters are directly related to the activation and modulation of host immune and inflammatory responses Table 1 Clusters 3 and 4 contain genes that are asso ciated with equally strong Sx response but responded discordantly Cluster 3 is denoted as LA BZ where superscripts nc and up stands for no change and upregulat
36. ificance of distributional differences Fisher s exact test lt 0 05 lt 0 005 doi 10 1371 journal pgen 1002234 9003 pro inflammatory and pro oxidant cytokine interleukin 1 beta IL1B 35 36 The NLR related genes are among the most highly differentially expressed genes discovered in our study These genes appear in two clusters cluster 2 Aroa and cluster 3 4 5 exhibiting markedly different temporal patterns Figure 2 Resid ing in cluster 3 NOD1 RIPK2 and CASP1 showed no significant change in Asx subjects q value gt 0 01 EDGE but highly increased among Sx individuals g value lt 0 0001 EDGE Figure 5A Figure S2 On the other hand NOD2 NLPR3 and CASP5 are found in cluster 2 Their expression decreased in Asx but increased evidently in Sx Figure 5B Figure 2 In addition the expression level of ILIB cluster 2 was evidently suppressed in the Asx phenotype while activated in the Sx phenotype Figure 5C Given the importance of NOD2 and NLPR3 to the processing of IL1B the Asx specific lower expression of IL1B may be contributed directly to the similar downregulation patterns of NOD2 and NLRP3 This hypothesis is supported by a new study in which Nod2 deficient mice showed decreased levels of TNF and ILIB in PBMC 34 Of relevance to the phenotypically different expression dynamics of NLR mediated inflammasome activation an opposite trend is observed in two cluster 6 4 S genes that are related to early
37. ion respectively The subscript mid middle stage indicates the onset time of the change Cluster 3 is characterized by strong activation in Sx phenotype of genes responsible for antiviral and inflammatory responses Cluster 4 A Sd contains genes that are continuously down regulated in the Sx phenotype in contrast to nearly no change in the Asx phenotype On the other hand genes in clusters 2 and 6 are associated with strong but discordant responses in both Asx and Sx individuals indicating an active physiological response in Asx hosts Cluster 2 A Sn q includes genes exhibiting sustained decrease unique to the Asx phenotype from early time onward In Sx the expression of cluster 2 genes increases to peak level at the middle of challenge 45 69 hpi followed by a rescinding trend Cluster 6 AE ry Sta is populated by genes whose expression steadily increases in the asymptomatic phenotype over all time In contrast for the symptomatic subjects these genes exhibit a transient but significant decrease beginning at 29 hpi and return to baseline after 60 hpi Host transcription signatures are highly correlated with disease dynamics The eight clusters represent molecular signatures of unique and contrasting temporal dynamics We evaluated whether these signatures are related to symptom development by correlating the expression of these signatures against standardized clinical symptom scores 17 18 Both positive and negative corre
38. is activated 36 hours before peak symptom time An examination of the highest ranked genes in cluster 3 L n Si reveals strong activation of host antiviral defense program Table 1 These genes include several PRR genes such as Toll like receptor 7 TLR7 the RNA helicases RIG I and interferon induced with helicase C domain 1 IFIH1 genes encodes proteins that are key to innate immune responses by acting as viral RNA sensors 12 26 28 These are among the most statistically significant g value lt 0 0001 EDGE exhibiting dramatic increase of expression starting at 45 hpi in Sx hosts Figure 4B Figure S8 Previous studies have demonstrated that the downstream signaling triggered by these PRRs converge at TANK binding kinase 1 TBK1 resulting in direct phosphory lation of interferon regulatory factor 7 IRF7 29 Both TBK1 and IRF7 Figure S1 have similar expression dynamics and are found in cluster 3 In total cluster 3 contains 11 genes from the August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A A B Asx Sx hpi 0 108 0 108 hpi A ADA Cluster COMMD4 d DDB2 1 EIF6 HLA DPA1 CCR1 MX2 WARS SPTLC2 ADAR SIGLEC1 LY6E IFI44L sco2 1F127 CD1C RPS4X CHI3L1 EEF2 ADAMTS5 ORM1 RPGR MPZL1 TXNDC13 FGFR10P RPL3 RPS16 RPL22 LOC442171 RPS3 CX3CR1 CSFAR 5 7 PSAP 8 IGSF2 n228 TNFRSF1B ALDH5A1 FBXO9 PRDX2 opc1 HOOK1 n 450 n
39. lations were observed Figure 3C In particular cluster 3 4 S7 showed the strongest positive correlation with symptom scores p 0 77 followed by cluster 2 p 0 58 The temporal expression pattern of cluster 3 genes closely resembled the disease progression trajectory of each individual Sx subject It is noteworthy that luster 3 is most significantly enriched with 70 of the BLU factor genes p lt 0 05 Fisher s exact test This is in strong concordance with the BLU gene expression signature being highly correlated with temporal disease progression Figure 3A and 3B Furthermore 90 of acute respiratory viral signature genes are found in E PLoS Genetics www plosgenetics org cluster 3 Table S5 17 In comparison the lack of symptoms in Asx subjects was consistent with their nearly constant low level expression of this same cluster of genes Figure 3B Interestingly the two largest clusters cluster 4 ne Sw and cluster 6 Arari Sie were the most negatively correlated with the development of symptoms p 0 54 and p 0 41 respectively Figure 30 These demonstrate the close association between the host tran scriptional signatures and the overt clinical disease development Cluster 6 contains significant proportion of genes encoding influenza virus interacting proteins A recent study identified 66 and 87 human proteins that physically interact with H3N2 Udorn and H1N1 A PR 8 34 PR8 viruses respec
40. liam L Sun B Rempel H 2004 Invasive chronic inflammatory monocyte phenotype in subjects with high HIV 1 viral load J Neuroimmunol 157 93 98 Chen G Shaw MH Kim YG Nunez G 2009 NOD like receptors role in innate immunity and inflammatory disease Annu Rev Pathol 4 365 398 Kanneganti TD Ozoren N Body Malapel M Amer A Park JH et al 2006 Bacterial RNA and small antiviral compounds activate caspase 1 through cryopyrin Nalp3 Nature 440 233 236 Kobayashi KS Chamaillard M Ogura Y Henegariu O Inohara N et al 2005 Nod2 dependent regulation of innate and adaptive immunity in the intestinal tract Science 307 731 734 Sabbah A Chang TH Harnack R Frohlich V Tominaga K et al 2009 Activation of innate immune antiviral responses by Nod2 Nat Immunol 10 1073 1080 Martinon F Burns K Tschopp J 2002 The inflammasome a molecular platform triggering activation of inflammatory caspases and processing of proIL beta Mol Cell 10 417 426 Martinon F Tschopp J 2005 NLRs join TLRs as innate sensors of pathogens Trends Immunol 26 447 454 Durand E Al Haj Zen A Addad F Brasselet C Caligiuri G et al 2005 Adenovirus mediated gene transfer of superoxide dismutase and catalase decreases restenosis after balloon angioplasty J Vasc Res 42 255 265 Oda T Akaike T Hamamoto T Suzuki F Hirano T et al 1989 Oxygen radicals in influenza induced pathogenesis and treatment with pyran polymer conjugated SOD Science 244
41. ly The resulted August 2011 Volume 7 Issue 8 e1002234 p values were adjusted for multiple comparisons with Benjamini Hochberg method 70 To identify canonical gene pathways in each SOM cluster that are highly associated with disease phenotypes we applied Globaltest 71 using the pathway definition in MsigDB database v2 5 72 that include both pathway components and targets We assessed the correlation between clinically determined symptom scores and the temporal gene expression of SOM clusters using standard linear mixed model regression The correlation R value was estimated using a signed coefficient of determination 67 68 The BLU factor analysis was used to detect disease signature shown in Figure 1A Unlike our implementation of EDGE SOM and Globaltest BLU is an unsupervised method requiring no prior class information Like other unsupervised Bayesian factor analysis methods BLU finds a decomposition of the data matrix Y here a p by n matrix of abundances of the p mRNA transcripts for each of m gene expression profiles into a matrix product MA where each column of M is a factor and each column of A is a set of factor loadings corresponding to individual factors in M for a given chip Y MA N In essence BLU estimates two matrix valued latent variables M and A whose product best approximates the most important information contained in the observation Y while minimizing the residual model fitting error denoted a
42. ly important is to understand the complexity of the host response in individuals who are exposed but effectively contain the infection and avoid symptomatic disease This study presents key transcriptional differences between Asx and Sx host responses and highlights an active state on a gene transcription level of viral control in both Sx and Asx hosts Viral sensing inflammation and symptomatic disease We showed that the viral sensing and inflammation in Sx hosts clearly correlate to clinical symptom development over time As mounting evidence has established the role of various PRRs in sensing viral components of influenza viruses our results confirm the concurrent activation of all known classes of PRRs and their signaling cascades by influenza viruses in human challenge models In contrast Asx hosts showed not only an absence of such activation but also negative regulation of related inflamma tory signals especially in the case of NLRP3 and NOD2 This corresponds to their lack of clinical apparent symptoms It has long been postulated that multiple PRRs represent a functional redundancy of host defense and that there exists signaling crosstalk among them stimulating similar cytokine profiles that are both pro inflammatory and pro oxidant 36 Here we found simultaneous and continued activation of all known PRRSs in Sx hosts with particular emphasis on NLR family genes Of important relevance two recent studies showed that HIN1 1918
43. matory branch of TLR signaling e g the interferon regu latory factor 5 IRF5 As a master regulator of the inflammatory arm of TLR7 signaling 9 IRF5 directly activates proinflamma tory cytokine tumor necrosis factor alpha TNF which has been directly implicated in flu like symptoms in many types of diseases with excessive inflammation These and other mediators of inflammatory response such as IL15 and IL10 genes share similar Asx specific increasing pattern Figure 4D Figure S7 Of interest the sialic acid binding Ig like lectin 1 SIGLEC1 or Sialoadhesin was strongly activated in Sx hosts at mid to late stage of infection Figure 4D As a macrophage specific adhesion molecule SIGLECI has recently been related to pro inflammatory function of macrophages in HIV infections 30 These results show that the expression kinetics of cluster 3 genes constitutes a transcrip tional signature of host antiviral program This signature fully presents itself 36 hours before the peak symptom time and it is indicative of disease severity Moreover its activation intensity maintained high level through 108 hpi An active asymptomatic state is characterized by down regulated expression of the NLRP3 inflammasome CASP5 and the IL1B pathway Members of cytoplasmic Nod NACHT LRR NLR family have recently been linked to pathogen pattern recognition Ori ginally identified in bacterial infections this family of molecules is important to the func
44. n This work represents by far the most extensive im vivo human challenge study on influenza viruses Combined with key clinical parameters our results offer an opportunity to look beyond individual signaling events and into their collective effects on symptomatic disease pathogenicity The detailed timing of various August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A A p value H060 3e 04 y 9 0 001 3 N 0 003 8 Ke o H 0 01 2 Wi A O E s 0 03 T EEEE E TT AAA O oi ke 0 3 6 j 1 12 12 36 60 84 108 12 36 60 84 108 Time hpi B H060 p value Uv E TT CT MI gt VD z 2 Ou a S E D ke N n Pl Jer Si E H S H 5 lt H E o OS m5 lt 12 12 36 60 84 108 12 36 60 84 108 Oo o E x a e nA A Ce Time hpi C Ribosomal Proteins Cluster 6 12 PBMC Lymphocyte 11 M 4 15 Signature 10 SOM 3 2 9 SOM 2 2 DA gs 3 SOM 1 4 7 7 A 6 Ka 5 014 BR TC 2 1 Cluster 6 PP E Ribosomal 12 0 5 12 21 29 36 45 53 60 69 77 84 93 101108 Protein Synthesis Time hpi Figure 6 Asymptomatic hosts showed unique temporal expression kinetics of cluster 6 genes related to JAK STAT signaling transduction and protein biosynthesis A B Distinct expression pattern of gene members in JAK STAT pathway and their association with symptom severity A Significant positive association between genes and disease severity is shown f
45. ng pattern Table SOA does not differ across inoculation dosage groups All nine symptomatic and four asymptomatic subjects showed shedding 21 25 Two asymptom atic shedders 14 and 16 are in lowest dosage group and the other two 2 and 4 are in the highest dosage group The amount of viral shedding are determined from nasal wash obtained daily Supplementary Methods Shedding values lt 1 25 are set to 1 25 in the plot PDF Figure S14 Asymptomatic subjects demonstrated non passive transcriptional response program As an example we show a significant temporal expression decrease of the inflammasome related gene NLRP3 in eight individual asymptomatic subjects Each subpanel depicts the temporal expression of one individual asymptomatic subject The y axis is the log base 2 signal intensity of NLRP3 and the x axis is the time from 12 hpi to 108 hpi hour post inoculation A polynomial fitting of expression values solid line was fitted using LOESS model and significance of temporal trend was assessed with EDGE Subjects 3 and 17 never showed detectable amount of virus lt 1 25 in their nasal wash Table S2 PDF Figure S15 Serological conversion versus clinical symptom outcome and gene expression The RPL3 gene expression trajectories for Asx blue and Sx red are representative of SOM cluster 6 Legend at right gives the character encoding of E PLoS Genetics www plosgenetics org 15 Temporal Host Molecular Resp
46. onents of the host immune system are diverse and they interact in a complicated manner Owing to both technical and ethical difficulties it has not been practical to experimentally determine the full course of immune responses leading to severe symptoms in otherwise healthy human hosts Thus the time sequence and orchestration of host response events remain to be fully understood August 2011 Volume 7 Issue 8 e1002234 Author Summary The transcriptional responses of human hosts towards influenza viral pathogens are important for understanding virus mediated immunopathology Despite great advances gained through studies using model organisms the complete temporal host transcriptional responses in a natural human system are poorly understood In a human challenge study using live influenza H3N2 Wisconsin viruses we conducted a clinically uninformed unsuper vised factor analysis on gene expression profiles and established an ab initio molecular signature that strongly correlates to symptomatic clinical disease This is followed by the identification of 42 biomarkers whose expression patterns best differentiate early from late phases of infection In parallel a clinically informed supervised analysis revealed over stimulation of multiple viral sensing pathways in symptomatic hosts and linked their temporal trajectory with development of diverse clinical signs and symptoms The resultant inflammatory cytokine profiles were shown to
47. onses to Influenza A each subject along with their disease outcome blue Asx and red Sx and their serologic conversion outcome converted and not converted There is no significant relation between disease outcome and serological conversion p value of 0 27 according to likelihood ratio test of dependency between these two outcomes The two seroconverted asymptomatic individuals subject 2 and 3 are called out by orange arrows in the gene expression trajectory plot The RPL3 expression profiles of these two subjects are not significantly different from those of the other asymptomatic hosts PDF Figure 16 Schematic outline of analysis pipeline Unsuper vised no clinical phenotype information was used Supervised clinical phenotype was incorporated in analysis PDF Figure S17 Daily white blood cell counts show mild change less than 80 from baseline in Asx and Sx phenotypes PDF Figure 18 Expression heatmap of genes that are significantly differentially expressed between Asx and Sx Genes are identified using EDGE g value lt 0 01 and clustered with SOM The average expression are computed and normalized for each gene to have zero mean and unit standard deviation Within a cluster genes are shown in decreasing order of significance level PDF Table S1 A Subject Demographic and Clinical Characteristics of Viral Challenge Cohort B Detailed subject demographics PDF Table S2 Viral
48. or 60 hpi left temporal gene expression pattern E PLoS Genetics www plosgenetics org 12 August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A of suppressor of cytokine signaling 1 SOCS1 shows upregulation in symptomatic hosts B Significant negative association between genes and disease severity is shown for 60 hpi left temporal gene expression pattern of SOCS2 shows upregulation in Asx hosts versus downregulation in Sx hosts C Significance of negative association p value between ribosomal protein synthesis RPS related genes and overall disease severity Pie chart top left shows a high degree of enrichment of significant RPS genes in SOM cluster 6 which is characterized by a trend of upregulation in Asx hosts versus downregulation in symptomatic hosts over time D Proportion overlap between cluster 6 ribosomal protein synthesis genes and lymphocyte signature ribosomal proteins genes 49 doi 10 1371 journal pgen 1002234 g006 immune response events in vivo will advance our understanding of their biological and clinical relevance to influenza virus mediated disease progression Materials and Methods Human influenza viral challenges We performed a healthy volunteer dose ranging intranasal challenge with influenza A A Wisconsin 67 2005 H3N2 at Retroscreen Virology LTD Brentwood UK We enrolled 17 pre screened volunteers aged 18 to 45 years of age who provided informed cons
49. osts and upregulation in symptomatic subjects C SOM cluster 2 gene interleukin 1 beta IL1B shows E PLoS Genetics www plosgenetics org 10 August 2011 Volume 7 Issue 8 e1002234 Temporal Host Molecular Responses to Influenza A symptomatic specific upregulation versus Asx specific downregulation over time D SOM cluster 6 genes superoxide dismutase SOD1 shows upregulation versus downregulation in Asx and Sx hosts respectively The expression values are plotted on a log base 2 scale and all genes are significantly differentially expressed between Asx and Sx q value lt 0 0001 EDGE doi 10 1371 journal pgen 1002234 g005 and influenced by specific cellular context Their contrasting expression dynamics in Sx versus Asx points to potential benefit in controlling inflammation by regulating NLRP3 mediated inflam masome activation or other inflammatory responses 55 Link between anti oxidant response and Asx infection The inflammasome and pro inflammatory cytokines have been linked to increased level of oxidative stress during viral infection 56 58 A recent report showed in mouse model that Nlrp3 inflammasome activation depends on reactive oxygen species ROS and inhibition of ROS induction abolished IL1B production during influenza infection 59 It is intriguing that our data shows a temporal Asx specific upregulation versus Sx specific suppression of SOD1 and SOK1 This coincides with the observed negative correlation be
50. performed using the Human Genome Ul33A 2 0 Array Affymetrix Santa Clara CA and expression profiles were pre processed using robust multi array RMA method 66 Text 1 Both raw and normalized gene expression data are available at GEO GSE30550 Statistical analysis Temporal gene expression was analyzed using EDGE 20 on RMA normalized intensities A total of 5 076 genes were identified as most significantly differentially expression genes g value lt 0 01 between Asx and Sx Co clustering of the significant genes found by EDGE was performed using Self Organizing Map 21 Text S1 We estimated the correlation between disease symptom scores and temporal expression values of clusters using a standard linear mixed model 67 68 Specifically for each individual symptom measured we regressed the scores onto the expression value vector of each SOM cluster separately with a random effects term accounting for within subject temporal correlation Biological pathway enrichment analysis was performed using Ingenuity Pathway Analysis IPA We implemented the non parametric Jonckheere Terpstra JT method 69 to test monotonicity of the expression patterns of individual gene clusters Briefly the JT test was applied independently to each cluster and configured to test the null hypothesis that there exists no monotonic trend in the temporal change of gene expression This test was performed separately for each one of two phenotypes separate
51. pgen 1002234 g004 immune cells to cytokine signals 40 Using pathway analysis we detected significantly distinct JAK STAT signaling dynamics p value lt 0 05 Globaltest involving two different sets of SOCS genes The first set included SOCS1 and SOCS3 from cluster 2 ad ST A while the second group consists of SOCS2 and SOCS5 a cluster 6 Aari Sra The expression of SOCS1 and SOCS3 declines at early time points among Asx but strongly increases among Sx Figure 6A Figure S11 Growing evidence suggests that SOCS1 and SOCS3 are important inhibitory modu lators in limiting the inflammatory effect of interferon signaling during viral infection 41 42 Our data supports such a protective role of SOCS1 and SOCS3 given their much higher levels of expression during late infection phase 45 hpi onward Consistent with cluster 6 but contrasting with the cluster 2 expression pattern Figure 2 SOCS2 and SOCS5 exhibits expression dynamics that clearly differ from that of SOCS1 and SOCS3 Starting from the early infection stage 712 hpi SOCS2 and SOCS5 show marked increasing trend in Asx and this trend persists throughout the entire infection period Figure 6B Figure S11 In contrast their expression diminishes in Sx especially between 45 hpi and 69 hpi A recent study showed that the anti inflammatory actions of aspirin induced lipoxins depend upon the function of SOCS2 43 Highly expressed in lymphoid organs SOCS5 was hypothesized to b
52. poptosis A Significance p value of association between phenotypes and the whole group of genes at all time points and at time 45 hpi top left panel B Average temporal expression intensities are computed on subjects in Asx and Sx phenotypes and normalized to have zero mean and unit standard deviation PDF Figure S6 Symptomatic specific temporal downregulation of cluster 4 genes n 13 that are related to mitogen activated protein MAP kinase cascades A Significance p value of association between phenotypes and the whole group of genes at all time points and at time 45 hpi top left panel B Average temporal expression intensities were computed on subjects in Asx and Sx and normalized to have zero mean and unit standard deviation PDF Figure S7 Increased temporal expression of inflammatory response regulators cluster 3 interleukin 15 and interleukin 10 The expression intensities are plotted on a log base 2 scale PDF Figure S8 Temporal gene expression of cluster 3 gene cytoplasmic double strand viral RNA sensor IFIH1 interferon August 2011 Volume 7 Issue 8 e1002234 induced with helicase C domain 1 The expression intensities are plotted on a log base 2 scale PDF Figure S9 Temporal expression of interferon inducible anti viral genes from cluster 3 The expression intensities are plotted on a log base 2 scale PDF Figure S10 Temporal gene expression of cluster 6 gene serine threonin kinas
53. s N in the formula above with latent variable order selection according to an hierarchical Bayesian model However unlike other factor analysis BLU decomposes the data into relative proportions such that the columns of M and the columns of A are non negative and the columns of A sum to one Intuitively a BLU discovered factor can be viewed as a gene expression profile whose amplitudes represent the relative contribution of each gene present in that factor and the factor loadings are the proportions of these factors that are present in each chip Such positivity constraints aid in interpre tation and are natural in gene microarray analysis as the expression intensity measurements of genes are always non negative BLU was run on all genes on the expression array and extracted a total of three major BLU factors The factor scores of the samples were subsequently divided into two groups samples taken before inoculation pre inoculation samples and samples taken after moculation post inoculation samples We then tested for significant difference between the scores of the pre inoculation and post inoculation samples t test with p value less than 0 01 At this significance level only one of the factors passed this test the acute respiratory factor shown in Figure 1A Based on the score of this acute respiratory factor we quantitatively determine the four regions by a threshold criterion using the pre inoculation samples The threshold was
54. s is unlikely for two reasons First the maximum observed change in cell populations for both Asx and Sx hosts was no more than 80 from baseline Figure S17 Second the distribution of leukocyte subpopulations is not correlated with phenotype at baseline or throughout the time course of the study Table S6 Thus the dramatic changes in gene expression described here cannot be attributed greatly to cell population changes Another uncontrolled factor is that certain subjects may have come into the study with related preconditions While we cannot completely dismiss the possibility of previous exposure to other respiratory viruses all subjects were healthy and tested negative for H3N2 influenza antibody at pre inoculation time None of the volunteers had been vaccinated for any influenza virus in the previous 3 years Finally while we did not observe subject demographics such as age gender or ethnicity to be influential of final disease outcome Table S1b we cannot rule out the possibility of small sample bias We have been careful to provide statistical safeguards against model overfitting by reporting significance measures p values and g values with qualifying confidence intervals that are associated with our findings Concluding remarks To our knowledge this multi institutional collaborative study presents the first systematic analysis of the full temporal spectrum of pathogen elicited host responses during influenza viral infectio
55. samples taken 24 hours prior to inoculation with virus baseline immediately prior to inoculation pre challenge and at set intervals following challenge peripheral blood for serum peripheral blood for RNA PAXgene nasal wash for viral culture PCR urine and exhaled breath condensate Peripheral blood was taken at baseline then at 8 hour intervals for the initial 120 hours and then 24 hours for the remaining 2 days of the study For all challenge cohorts nasopharyngeal washes urine and exhaled breath condensates were taken at baseline and every 24 hours Samples were aliquoted and frozen at 80 C immediately RNA purification and microarray analysis RNA was extracted at Expression Analysis Durham NC from whole blood using the PAXgene 96 Blood RNA Kit PreAnalytiX Valencia CA employing the manufacturer s recommended protocol While whole blood RNA is initially extracted a secondary procedure B globin reduction was then employed to remove the contribution of red blood cell RBC RNA to the total RNA A set of four peptide nucleic acid PNA oligomers whose sequences are complementary to the 3 portions of the alpha and beta hemoglobin RNA transcripts were added to reduce globin RNA transcription due to RBC The inhibition of globin cDNA synthesis dramatically reduces the relative amount of anti sense biotin labeled cRNA corresponding to the hemoglobin transcripts Hybridization and microarray data collection was
56. set to be more than 4 times the maximum pre inoculation sample score corresponding to a test p value less than 0 05 Text S1 In this manner all samples were labeled with one of four classes namely classes 1 4 Figure 1A The class designation of a sample indicates distinct risk levels of four intrinsic disease states uninfected class 1 infected with low risk for symptom development class 2 infected with high risk for symptom development class 3 and infected with overt symptoms class 4 The genes exhibiting largest contrast between each pair of classes were extracted from all genes on the expression array using a LogitBoost classifier 73 as a contrast function Note that our objective is not to obtain a classifier between regions but rather to use LogitBoost to identify groups of genes most associated with E PLoS Genetics www plosgenetics org 14 Temporal Host Molecular Responses to Influenza A differences between a pair of classes As it uses boosting algorithm to perform variable selection our implementation of LogitBoost yields a set of genes in addition to a classifier function To do this we generated 200 bootstrap samples from each class 74 We randomly selected 2 3 of each bootstrap sample to construct the boosting ensemble and the other 1 3 of data was used to evaluate the variability of the association between the largest contrast genes and each class pair We defined the largest contrast genes as the set
57. tion of innate immunity 31 33 A recent study showed that nucleotide binding oligomerization domain 2 NOD2 recognizes ssRNA of both Influenza and respiratory syncytial viruses 34 Furthermore activated NODs were linked to the activation of receptor interacting serine threonine kinase 2 RIPK2 and subsequently nuclear factor kappa B NFkB activation whereas activated NLPRs result in forming so called inflammasome complexes This process involves caspase 1 CASP1 and caspase 5 CASP5 and ultimately the release of Table 1 Canonical pathways and representative genes enriched in individual SOM clusters SOM Cluster of Genes Pathway Representative Genes 1 450 immune cell trafficking antigen presentation CD74 HLA DMA HLA DPA1 HLA DPB1 CCR5 CCL4 TBX21 IL10RA CD244 ICAM2 2 759 inflmmation chemotaxis of macrophage SOCS1 SOCS3 NOD2 NLRP3 CASP5 IL1B STAT3 ADM C5 CCL2 7 8 11 neutrophils and dendritic cells antigen CCR1 CCR4 CD14 CD59 CD163 CD209 CEACAM3 CXCL9 CXCL10 presentation JAK STAT signaling CXCL11 FAS HLA B ICAM1 IL17RB IL18R1 IL18RAP LILRA2 LTBR MX2 TGFB1 TLR1 TLR2 TLR4 TLR5 TLR8 TREM2 TRIM21 SERPINA1 CASP4 IFITM2 3 739 inflammatory response dendritic cell and TLR7 MYD88 IRF7 IRF5 IRF9 TNF JAK2 PSMB8 STAT1 DDX58 IFIH1 neutrophil activation IFN signaling IL18 IL10 MX1 RSAD2 OAS1 SIGLEC1 NOD1 CASP1 PKR TRIM22 LILRB1 ISG20 IFNAR1 IFI44 CD86 CD40 CD63 C1QA
58. tively 25 We examined the distribution of genes corresponding to these proteins among the eight clusters identified in our analysis Several interesting findings result from the comparison A total of 27 45 and 40 46 respectively of genes overlap with the set of differentially expressed genes found in our study Figure 3D The majority of these genes 67 are found in cluster 4 and 6 Except for clusters 2 and 3 the H3N2 Udorn and H1N1 PR8 genes are distributed in a similar proportion across the eight SOM clusters Such similarity shows functional conservation between the two viral strains Secondly cluster 6 alone contains 44 of the 27 overlapping genes H3N2 Udorn This is significantly disproportional to the size of cluster 6 p value lt 0 05 Fisher exact test Several of the overlapping genes such as PRKRA MAPK9 and NRFl have been shown to play important roles in host immune or antioxidant function Thirdly cluster 2 and 3 showed a significantly lower proportion of overlapping genes p value lt 0 05 Fisher s exact test These results suggest that genes in these two clusters are more likely to be indirectly regulated by the viruses such as those involved in inflammatory responses Taken together the results independently validate the functional relevance of the molecular signatures identified in our challenge study and suggest that many cluster 6 genes might be directly regulated by viruses The host antiviral program
59. tween these genes and NLRP3 Since SOD1 and SOKI are capable of reducing ROS and of suppressing oxidative stress 37 their increased expression in Asx hosts may play a role in negatively regulating NLRP3 expression and inflammasome signaling In support of this hypothesis is a study on the efficacy of antioxidant therapy found that pyran polymer conjugated SOD1 protected mice against potentially lethal influenza virus infections 38 Together our results provide evidence for a protective role of antioxidants SOD1 and SOKI1 Their increased mRNA expression may constitute an effective antiviral mechanism by which aberrant immune responses are avoided in Asx hosts The nature of Asx phenotype It is estimated that Asx infections account for 30 50 of seasonal flu cases 2 which is consistent with the attack rate in our study Since both Asx and Sx subjects were challenged under the same protocol and displayed inoculation dosage independent viral shedding this raises a critical question concerning the nature of the observed Asx phenotype We have strong evidence that the observed Asx molecular signatures are a consequence of rapid innate response rather than being due to failed inoculation Firstly 50 of Asx subjects had evident viral shedding This is on par with that of subclinical or secondary infections reported in the literature In addition serum neutralizing antibody nAb titre were nearly identical in Asx and Sx subjects
60. ud D Turner RB Winther B Wiehler S Tiesman JP et al 2008 Gene expression profiles during in vivo human rhinovirus infection insights into the host response Am J Respir Crit Care Med 178 962 968 Zaas AK Chen M Varkey J Veldman T Hero AO 3rd et al 2009 Gene expression signatures diagnose influenza and other symptomatic respiratory viral infections in humans Cell Host Microbe 6 207 217 Jackson GG Dowling HF Spiesman IG Boand AV 1958 Transmission of the common cold to volunteers under controlled conditions I The common cold as a clinical entity AMA Arch Intern Med 101 267 278 Dobigeon N Moussaoui S Coulon M Tourneret J Y Hero A 2009 Joint Bayesian endmember extraction and linear unmixing for hyperspectral imagery IEEE Trans on Signal Processing 57 4355 4368 Storey JD Xiao W Leek JT Tompkins RG Davis RW 2005 Significance analysis of time course microarray experiments Proc Natl Acad Sci US A 102 12837 12842 Kohonen T 1995 Self Organizing Maps Series in Information Sciences 30 B hlmann P Hothorn T 2007 Boosting algorithms regularization prediction and model fitting Statistical Science 22 477 505 Samuel CE 2001 Antiviral actions of interferons Clin Microbiol Rev 14 778 809 table of contents Manderson AP Botto M Walport MJ 2004 The role of complement in the development of systemic lupus erythematosus Annu Rev Immunol 22 431 456 Shapira SD Gat Viks I Shum BO Dricot A de Grace

Temporal Dynamics of Host Molecular Responses

Contents

Download Pdf Manuals

Related Search

Related Contents