Pseudogenes and neutral evolution in Drosophila melanogaster

Abstract

An understanding of the rates and patterns of neutral evolution is important for estimating divergence times and for recognising selection acting upon the genome. In order to characterise neutral evolution it is necessary to identify sequences that are evolving free from selective constraint. This has proven difficult in Drosophila, where sites such as synonymous sites, introns, and untranslated regions of genes have been shown to have selection acting upon them. Pseudogenes are inactive copies of genes that are by definition functionless, and are thus ideal candidates for the study of neutral evolution. Historically few pseudogenes have been known in Drosophila, and many sequences that have been thought to be pseudogenes have been found to be functional. The advent of the genomics era has allowed for the identification of many more potential pseudogenes. In this thesis I study these in order to identify likely genuine pseudogenes, and use these to characterise neutral evolution in Drosophila.

The first step of this project was to identify a list of genuine pseudogenes. Two pseudogene datasets were used: those identified in a paper by Harrison et al., (2003), and those listed on the Drosophila genome database Flybase. A number of techniques were used to study these pseudogenes, including cDNA analysis, conservation analysis, and resequencing. The results of this study showed that many of these sequences were not genuine pseudogenes and had been incorrectly assigned due to incorrect genome annotation, unknown splicing patterns, and polymorphic inactivating mutations. In total 73 likely Drosophila melanogaster pseudogenes were identified.

The technique Gene Identification by Nonsense-Mediated Decay (GINI) (Noensie & Dietz, 2001) was investigated as a possible technique for identifying new pseudogenes. This involved feeding Drosophila larvae drugs known to inhibit nonsense-mediated decay and then determining whether this resulted in an upregulation of transcripts known to contain premature termination codons (PTCs). Caffeine was found to result in the upregulation of 4 of 6 PTC-containing transcripts, most notably those with longer 3’ UTRs. This technique thus seems promising for the identification of new pseudogenes.

Following this, 47 pseudogenes were resequenced in three populations. It was found that ancient pseudogenes had higher levels of nucleotide diversity than recently inactivated genes and new pseudogenes, presumably due to their time since inactivation and pseudogene population size during this time. Various aspects of pseudogene evolution were characterised including nucleotide diversity, mutation patterns, and FST, allowing us to gain a better understanding of population structure and the background mutational patterns of the genome.

Finally, I investigated several hypotheses as to why the Esterase-7 gene might have a large proportion of inactivated alleles in natural populations of D. melanogaster. I concluded that the selective constraint acting on Esterase-7 has likely been relaxed in D. melanogaster relative to other species, and Esterase-7 may be in the process of becoming a pseudogene. This gives us key insights into the pseudogenisation of functional genes by mutation and genetic drift - mutations gradually occur and spread throughout the population while functional copies still remain in the population.