Evolutionary Systems Biology Seminars

Next week on Thursday (17th April 2008 ) the University of Edinburgh will host two seminars on evolutionary systems biology by Prof. Laurence Hurst from the University of Bath (homepage):

• Splicing and the evolution of protein coding genes
  Institute of Evolutionary Biology, Genetics Journal Club
  1 pm, Lecture Theatre 3, Ashworth Labs, Kings buildings, University of Edinburgh.
• Why isn't gene order random in eukaryotic genomes?
  Center for Systems Biology Edinburgh Seminar Series
  4 pm, Lecture Theatre 1 (G.27), Rutherford Bldg, Kings Buildings
  

The seminars were organized by Laurence Loewe (contact) with the help of the Centre for Systems Biology Edinburgh and will highlight the potential of combining systems biology approaches and evolutionary theory in the analysis of genomic sequences and structures (see the abstracts below). The corresponding work belongs to the finest recent examples of how evolutionary theory helps understanding biology.

 

Splicing and the evolution of protein coding genes

Prof. Laurence D. Hurst (University of Bath)
1:00 pm (17 April 2008), Lecture Theatre 3, Ashworth Labs
Institute of Evolutionary Biology (events), Genetic Journal Club
Kings Buildings, University of Edinburgh, UK

Until recently it has been considered that the sequence information in pre-mRNA necessary for accurate removal of introns was almost exclusively in the introns.  Over the last decade this position has changed.  In many species, mammals included, motifs in exons are necessary for intron removal.  Chief amongst these motifs are exonic splice enhancers (ESEs), A rich C poor motifs, necessary for the binding of SR proteins which in turn direct accurate splicing.  That sequences in exons have two roles - to specify codons and to direct splicing - suggests unrecognized dual-coding constraints might exist in the evolution of mammalian genes.  In this talk I show that the preservation of ESEs impacts on synonymous rates of evolution and codon bias, providing evidence for common selection on synonymous mutations, previously considered an unorthodox possibility in mammals. 
In Drosophila this splice related codon bias is also opposed to that preferred for translational optimization and explains unusual codon bias at exon ends. ESEs also skew amino acid usage and non-synonymous rates of evolution. Indeed, as predictors of rates of evolution dual coding constraints are one of the best if not the best predictor. These skews are seen in species which a priori are expected to have problems recognizing exons, ie those with relatively abundant intronic DNA.

Why isn't gene order random in eukaryotic genomes?

Prof. Laurence D. Hurst (University of Bath)
4:00 pm (17 April 2008), Rutherford Building Lecture Theatre 1 (G.27)
Centre for Systems Biology Edinburgh Seminar (link)
Kings Buildings, University of Edinburgh, UK

For a few years now it has been recognized that gene order in eukaryotes is not random. In particular, genes of comparable expression profile tend to cluster in genomes. In yeast, for example, highly co-expressed genes tend to cluster, while in mammals broadly expressed genes cluster. These findings suggest several new problems. First, is this clustering the result of natural selection or might it simply reflect leaky expression of neighbouring genes? Second, does clustering for co-expression explain all clustering? I shall present evidence that selection does act on gene order to maintain physical proximity of co-expressed genes. However, I shall also show that linkage per se ensures some degree of co-expression (even controlling for the similarity in transcription factor control) and only for the most highly co-expressed gene pairs need we suspect the action of selection. As regards the final issue, I argue that selection for minimization of noise in gene expression can explain many of the unusual features of the distribution of genes in the yeast genome, not least of which is the clustering of essential genes.