FROM MISSING BIRD GENES TO ACCELERATED EVOLUTION IN VERTEBRATES USING A GENOME-WIDE LANDSCAPE APPROACH

During this PhD project, a new visualization method was developed to assess regional effects of base composition and protein evolution in vertebrate genomes. Data of neighboring genes were integrated into genome-wide landscapes with information about regions on the chromosome rather than about individual genes.We applied this new approach to further study the enigma of "missing" avian genes. More than thousand genes are currently unaccounted for in bird genomes: too many to make sense from a biological point of view. This raises the question whether these genes were really evolutionary lost or whether they still need to be discovered. Our genome landscape approach shows that these "missing" genes are concentrated in more than 10 clusters of around 100 genes of the human reference genome. In these "missing gene clusters", the majority of the genes can be found in at least one avian species, indicating they should also be present in other avian species. Those gene sequences are often partial sequences with a high amount of G and C bases. This high GC content hinders the sequencing process, leading to the artefact of missing genes. One example is the gene encoding for glucose transporter 4 (GLUT4), which is important for glucose homeostasis in vertebrates, as it is insulin-regulated and highly expressed in skeletal muscle. It was believed for decades that GLUT4 is truly "missing" in birds which are hyperglycemic according to human standards and insulin resistant. However, during this project, we have sequenced the chicken transporter and demonstrate that the gene is expressed in skeletal muscle, both on the mRNA and protein level.The location of "missing genes" in regions where the genes have high GC content and accelerated protein evolution is linked to a genetic mechanism: GC biased gene conversion. This process occurs during meiotic recombination, where mismatches are repaired with a slight bias for GC alleles. The rate of GC biased gene conversion is proportional to the recombination rate and therefore highest in subtelomeric chromosome regions. Interestingly, bird genomes have microchromosomes in which all genes can be considered as subtelomeric. We noted that even in the best characterized avian genomes some microchromosomes are not yet annotated.A second application of our landscape approach was a comparative analysis of the genomes of Eutheria (placental mammals) and Metatheria (marsupials). We identified areas of accelerated protein evolution that coincided with regions where the genes accumulate GC bases. As predicted by the model of GC biased gene conversion, such regions are most often located in subtelomeres of the human and opossum reference genomes. Interestingly, we observed landscapes with lineage-specific subtelomeric effects of GC% and protein divergence in both eutherian and metatherian genomes. The divergent accelerated evolution of specific Eutherian and Metatherian subtelomeres is proposed to have contributed to divergence of reproductive physiology and adaptive immunity in these two mammalian taxa.We anticipate that the genome-wide landscape approach that was investigated in this thesis can be applied to other studies of genome evolution and genome biology, not only in vertebrates but in a wide range of other species.

Jaar van publicatie:2020