E-GEOD-35888 - Mapping Clinical and Expression QTL in a Sex-Dependent Effect of Host Susceptibility to Influenza H3N2/HK/1/68-MA20

Status
Released on 1 May 2012, last updated on 27 June 2012
Organism
Mus musculus
Samples (54)
Array (1)
Protocols (6)
Description
Seasonal influenza outbreaks and recurrent influenza pandemics present major challenges to public health. By studying immunological responses to influenza in different host species, it may be possible to discover common mechanisms of susceptibility in response to various influenza strains. This could lead to novel therapeutic targets with wide clinical application. Using a mouse-adapted strain of influenza (A/HK/1/68-MA20 [H3N2]), we produced a mouse model of severe influenza (p-flu) that reproduces the hallmark high viral load and overexpression of cytokines associated with susceptibility to p-flu in humans. We mapped genetic determinants of the host response using a panel of 29 closely related mouse strains (AcB/BcA panel of recombinant congenic strains) created from influenza-susceptible A/J and influenza-resistant C57BL/6J (B6) mice. Combined clinical quantitative trait loci (cQTL) and lung expression QTL (eQTL) mapping identified candidate genes for two sex-specific QTLs on chromosomes 2 and 17. The former includes the previously described Hc gene, a deficit of which is associated with the susceptibility phenotype in females. The latter includes the phospholipase gene Pla2g7 and Tnfrsf21, a member of the tumor necrosis factor receptor superfamily. Confirmation of the gene underlying the chromosome 17 QTL may reveal new strategies for influenza treatment. To identify eQTLs, we reanalyzed lung expression data previously obtained by Lee et al (2006) on MGU74Av2 microarrays (Affymetrix) for 54 mice (13BcA, 12AcB, B6 and A/J mice in duplicate) using Custom CDFv12 to reorganize oligonucleotide probes based on the latest genome and transcriptome information. We inferred an A/J or B6 strain of origin for each gene based on the genotype of surrounding markers with a call rate of 96.7%. If a gene was in between A/J and B6 markers for a given RCS, it was coded as NA. Expression values were normalized using the Robust Multi-array Analysis (RMA) for Affymetrix gene chips. To define the association between differentially expressed genes and genotypes, ANOVA was conducted on a per-gene basis using the linear model Expression ~ DSO + BG + DSO*BG, where background (BG) and donor strain of origin (DSO) were coded as binary phenotypes corresponding to A/J or B6. The cutoff for genome-wide significance was computed using the Benjamini–Hochberg correction. For reference, the paper by Lee (2006): http://www.ncbi.nlm.nih.gov/pubmed/16449383
Experiment type
transcription profiling by array 
Contacts
Rob Sladek <rob.sladek@mcgill.ca>, Anny Fortin, Bing Ge, Celia M Greenwood, Donna Sinnett, Earl G Brown, Emil Skamene, Gregory A Boivin, J C Loredo-Osti, Julien Pothlichet, Marina Takane, Michael Hallett, Peter D Lee, Robert Sladek, Sebastien Brunet, Silvia M Vidal, Thomas J Hudson, Tomi Pastinen, Yannick Fortin
Citation
MIAME
PlatformsProtocolsFactorsProcessedRaw
Files
Investigation descriptionE-GEOD-35888.idf.txt
Sample and data relationshipE-GEOD-35888.sdrf.txt
Raw data (1)E-GEOD-35888.raw.1.zip
Processed data (1)E-GEOD-35888.processed.1.zip
Array designA-AFFY-6.adf.txt
R ExpressionSetE-GEOD-35888.eSet.r
Links