De-novo transcriptome sequencing of a normalized cDNA pool from influenza infected ferrets

Abstract

The ferret is commonly used as a model for studies of infectious diseases. The genomic sequence of this animal model is not yet characterized, and only a limited number of fully annotated cDNAs are currently available in GenBank. The majority of genes involved in innate or adaptive immune response are still lacking, restricting molecular genetic analysis of host response in the ferret model. To enable de novo identification of transcriptionally active ferret genes in response to infection, we performed de-novo transcriptome sequencing of animals infected with H1N1 A/California/07/2009. We also included splenocytes induced with bacterial lipopolysaccharide to allow for identification of transcripts specifically induced by gram-negative bacteria. We pooled and normalized the cDNA library in order to delimit the risk of sequencing only highly expressed genes. While normalization of the cDNA library removes the possibility of assessing expression changes between individual animals, it has been shown to increase identification of low abundant transcripts. In this study, we identified more than 19,000 partial ferret transcripts, including more than 1000 gene orthologs known to be involved in the innate and the adaptive immune response.

Figure 1. The histogram in panel A shows the distribution of contigs length of the sequence assemblies.

The red bars indicate the length of the contigs from the assembly using the sequences from the cDNA library from infected tissues. The green bars show the contigs lengths for the combined assembly, built from these sequences as well as the ESTs previously reported . The histogram in panel B illustrates the number genes that were represented by one or multiple non-overlapping of contigs in the combined assembly. Panel C shows the distribution of species extracted using the best hit (lowest blast E-value) from the Refseq Protein blast analysis of the combined assembly.