Computational and Statistical analyses of Molecular Evolution and Demography using Large-scale Sequencing Data

Evolution can be described as the change of allele frequencies over time. Four forces - mutation, migration, genetic drift, and selection, drive this change. The aim of my thesis was to accurately estimate and differentiate the parameters governing each of these four mechanisms by utilizing various types of Next-Generation Sequencing datasets. More specifically, in chapters 1 and 2, I focused on investigating how the past demographic history of African and European D. melanogaster affected its genomic polymorphism. Modern genomes of flies carry signatures of past events such as migration to new regions, adaptation to new environments, and population size changes. By studying whole genome sequences of 29 wild strains from West Africa, 14 from Sweden and comparing them with genomes from Zambia (the putative ancestral range of the species), we were able to report for the first time, colonization time of the Western part of the African continent at approximately 72k years ago. Additionally, we demonstrated the importance of gene flow between the two populations, as well as, current and past effective population sizes. Our estimations confirmed already published predictions (Current Zambian and Swedish population size, ancestral African population size). Finally, we demonstrated the importance of inversions when accounting for demographic events of D. melanogaster. In chapter 3 of my thesis, I evaluated the importance of selection acting on the DNA-binding residues of the biggest family of transcription factors in the primates, namely KRAB-ZF genes. We were able to demonstrate the existence of two distinct sub-groups, based on the type of polymorphism (synonymous or not) carried by the DNA-contacting nucleotides. The two groups of genes differ by their expression breadth and intensity, as well as at the number of paralogs and orthologs and their evolutionary age. Additionally, we manually annotated the complete catalog of human KRAB-ZF genes, thereby providing a valuable resource for further investigation of this family of genes. In conclusion, the work carried during my thesis enabled to refine the evolution and demography of D. melanogaster African and Northern European populations, underlying the importance of modelling migration flows between populations for accurate estimation of split time. The second component of my thesis demonstrated the applicability of transcriptomics and epigenomics datasets to study evolution of the KRAB-ZF family. The proposed methodologies are applicable to other transcription factor gene families and our manually curated dataset is relevant to other scientists deciphering the function of these genes.

Elucidating regulatory mechanisms shaping the transcriptome and proteome of the gut immune response in Drosophila melanogaster

Michael Vincent Frochaux

Work on Drosophila melanogaster paved the way to our current understanding of modern genetics. Since then, this model organism has contributed greatly to various fields such as neurobiology, development, and immunology. The discovery and analysis of the various pathways of the Drosophila immune response led the way to the in-depth characterization of the gut immune response. Furthermore, a genome-wide association study on a population of Drosophila identified resistance to infection by the entomopathogenic bacteria Pseudomonas entomophila (P.e.) as a complex phenotype, with highly resistant and susceptible lines, and highlighted genes and pathways mediating inter-individual differences. However, the molecular mechanisms mediating the resistance to enteric infection remain largely uncharacterized. This study uses the same Drosophila population to analyze what are the molecular mechanisms mediating the resistance to enteric infection. We analyzed the genetic determinants of gene expression variation among the most resistant and susceptible lines in response to infection using expression quantitative trait loci (eQTL) analysis. We also characterized the mode of action of these eQTLs into cis- and trans-acting. Furthermore, we identified the gene nutcracker (ntc) as the only differentially-expressed gene between resistance classes. We then demonstrate that loss of function ntc mutants are significantly more susceptible to infection and then identify a single nucleotide polymorphism (SNP) located in a transcription factor binding site (TFBS) which affects the binding affinity of the transcription factor Broad leading to allele-specific expression variation of the gene ntc. If the transcriptomic response to infection has been thoroughly characterized, it is not the case for the proteomic response. We thus sought to characterize for the first time the proteomic response of the Drosophila to enteric infection. We performed mass spectrometry and RNA sequencing on dissected guts from flies infected by two Gram-negative bacteria, Erwinia carotovora carotovora 15 (Ecc15) or Pseudomonas entomophila, 4h and 16h after infection. We found that a large portion of the measurable proteome (12%) varies after infection and that protein changes are strongly time- and infection-dependent. We showed the relatively poor correlation between gene expression and protein abundance in the Drosophila enteric immune response. Finally, we performed a screen of several proteins that were identified in our proteomics work but had previously not been found when assessing the gene expression response to infection using loss-of-function mutants and identifying 7 genes that modulate overall susceptibility to P.e. infection. In summary, this work analyze the genetic determinants of gene expression variation in the DGRP population upon infection and characterize them according to their mode of action. Furthermore, we describe how a non-coding variant lowers resistance to infection by modulating ntc gene expression through altered Broad repressor binding. Finally, it provides the first comprehensive characterization of the Drosophila gut proteome upon infection by Gram-negative bacteria which can be the basis of future proteomic analysis of the enteric immune response

EPFL2019

The evolution of gene expression and binding specificity of the largest transcription factor family in primates

Jeffrey David Jensen, Adamantia Kapopoulou, Lisha Mathew, Didier Trono

The KRAB-containing zinc finger (KRAB-ZF) proteins represent the largest family of transcription factors (TFs) in humans, yet for the great majority, their function and specific genomic target remain unknown. However, it has been shown that a large fraction of these genes arose from segmental duplications, and that they have expanded in gene and zinc finger number throughout vertebrate evolution. To determine whether this expansion is linked to selective pressures acting on different domains, we have manually curated all KRAB-ZF genes present in the human genome together with their orthologous genes in three closely related species and assessed the evolutionary forces acting at the sequence level as well as on their expression profiles. We provide evidence that KRAB-ZFs can be separated into two categories according to the polymorphism present in their DNA-contacting residues. Those carrying a nonsynonymous single nucleotide polymorphism (SNP) in their DNA-contacting amino acids exhibit significantly reduced expression in all tissues, have emerged in a recent lineage, and seem to be less strongly constrained evolutionarily than those without such a polymorphism. This work provides evidence for a link between age of the TF, as well as polymorphism in their DNA-contacting residues and expression levels-both of which may be jointly affected by selection.

Wiley-Blackwell2016

Elucidating regulatory mechanisms shaping the transcriptome and proteome of the gut immune response in Drosophila melanogaster

Michael Vincent Frochaux

EPFL2019