Quantitative proteomics on the fly

The development of multicellular organisms is characterized by complex processes that progressively transform essentially a single cell into a creature with complicated structures and highly specialized functions. The fruit fly Drosophila melanogaster provides an excellent model system to investigate these principles of life in great detail. Over the past few decades, Drosophila has been used to investigate and elucidate fundamental aspects that underlie the mechanisms described above. Since this research mainly involved large-scale analysis of genes, many principles are now understood at the gene level. However, it is nowadays clear that transcript abundances no not necessarily correlate with protein expression levels and since the latter determine the complexity of cells it is the ultimate goal to investigate and understand these processes directly at the protein level. The maturation of proteomics-based mass spectrometric techniques allows capturing the identity of many proteins in a single experiment. The primary goal of the work described in this thesis focused on gaining insights into the dynamics of early embryonic development of the fruit fly Drosophila melanogaster at the protein level by combining stable isotope labeling with high-accuracy mass spectrometry. We have developed a quantitative proteomic approach utilizing in vivo 15N-labeling of fruit flies with stable isotopes combined with extensive analysis by LC-MS/MS which has permitted the relative quantitation of thousands of proteins during early embryonic development. This provided insight into the production, stability and modification of individual proteins, while discrepancies between transcriptional profiles and protein dynamics indicated novel control mechanisms in genome activation during early fly development. Differential regulation of numerous proteins was observed and allowed the classification of these proteins into specialized classes. Although this approach shows its strength in identifying and quantitating proteins, some drawbacks include suboptimal labeling and lack of appropriate software for data processing. To overcome these issues, alternative objectives were explored in the work described in this thesis and aimed at developing methods to optimize, improve or facilitate the quantitative analysis of multiple proteins in a single experiment. Metabolic labeling with suboptimal (<98%) 15N-enrichments negatively affects protein identification and quantitation which was discovered by the systematic investigation of two independent 15N-labeled datasets. To overcome or compensate these shortcomings we have developed methods that can be applied to qualitative and quantitative 15N-labeled data. Although metabolic labeling produces most accurate quantitative data, it is not always feasible to incorporate an internal standard metabolically into an organism. So-called label-free techniques allow the quantitation of proteins in two or more samples without the need of using an internal standard. By directly comparing protein expression levels obtained by metabolic labeling based on stable isotope 15N-labeling and label-free quantitation based on spectral peak areas we have correlated both approaches and preliminary data suggests that there is a weak positive correlation between these approaches. Taken together, we have presented a powerful comprehensive proteomics approach that we have applied to early embryonic developed of the fruit fly which provided insights into the production, stability and modification of individual proteins during the developing fly.