Centrifugation based approaches for isolation of the nuclear and the cytosolic proteomes were challenging to obtain pure fractions in cultured cells. Hence, I shifted to a different model organism, X. laevis, from which I can obtain the cytosolic and the nuclear proteomes with unmatched purity. Absolute amounts of ~4400, ~4800 and ~5500 proteins in the cytosolic, nuclear, and total oocyte extracts, respectively, were estimated with spiking of protein standards with known amounts (UPS2, Sigma). In total, proteomics of X. laevis oocyte provided MS-based quantitative information for ~6300 proteins. This high number of identification level by was achieved by employing molecular weight based separation at the protein level and followed by trypsin digestion of proteins.
A drawback of X. laevis model system is the lack of well-annotated genomic data. This dramatically hampers proteomics studies, which requires well-annotated genomic sequence for faithful protein identification. In order to increase the identification level, generated MS data were searched not only against X. laevis database and but also against an evolutionarily close species X. tropicalis. During the preparation of this thesis, a study compiled an mRNA derived protein reference database for X. laevis and based on this database, authors reported identification of ~11000 proteins from X. laevis eggs (Wühr et al., 2014). This indicates that my spatial proteome analysis of X. laevis oocyte is not at the saturation level for protein identification. This could be due to the two reasons. First, their mRNA derived reference database is more comprehensive protein sequence database than the one I used. Authors
laevis protein databases. Initial testing of this mRNA-based protein database has resulted in 15%
more protein identification for my data set (data not shown). Next, it appears that I could not reach the complete depth for the protein identification. Complementary techniques can be employed to increase the level of protein identification. Instead of a fractionation at protein-level, a peptide level fractionation technique such as peptide isoelectric focusing, size exclusion or ion exchange chromatography can be employed. Alternatively, protein digestion with different endopeptidase can be performed to generate differential peptide spectrum for MS detection. An unavoidable drawback of X. laevis model system is the presence of high number allelic protein variants as a result of pseudotetraploid genome. This leads to sequencing of more peptides species from same protein, therefore decreasing the proteins identification level.
Quantitative analysis of X. laevis proteome revealed interesting observations. Only a handful of proteins dominate majority of the protein molecules in the oocyte. The 22 most abundant proteins constitute 25 % of the total protein molecules in the oocyte. A striking example is, actin, which is the most abundant protein it comprising ~2% of the total protein molecules. Additionally, ribosomal proteins cover almost 20% of the all protein molecules in the oocyte. These observations underline the fact that X. laevis oocyte spends most of its resources to produce many copies of a few proteins, similar to yeast and humans cells (Marguerat et al. 2012; Nagaraj et al. 2011).
The most abundant proteins in the total oocyte extract belong to ribosomal proteins, chaperones, metabolic pathway members, cytoskeleton proteins, and several RNA binding proteins. In general, ribosomes, metabolic pathway members, and cytoskeleton proteins are highly abundant across all organisms and cell types (Marguerat et al. 2012 Nagaraj et al. 2011). Still, presence of huge amounts of ribosomal proteins, metabolic pathway members may reflect oocyte’s unique physiological needs. During fertilization, apart from equal number of chromosome contribution from sperm and egg, egg provides majority of the biological material, which will be needed in early embryonic development. Therefore, oocyte stockpiles huge amounts of biological material (protein, RNA, lipids) for later usage. Even though, oocyte is quite active in protein synthesis during development, it has been reported that only a portion of the ribosome is active and they are stored for later use (Taylor et al., 1985). Metabolic pathway members, particularly, glycolysis members might be needed to contribute energy requirement of the giant oocyte during oogenesis and later during embryogenesis. Highly abundant cytoskeleton proteins (several of them are actin, cofilin, gelsolin) are needed for mechanical stability of the giant oocyte.
After fertilization, egg will undergo 12 rounds of cell division without G1 phase. During these cell divisions, fertilized egg is practically silent in transcription and uses stored mRNAs for new protein synthesis (Roeder, 1974; Newport & Kirschner, 1982). The comparison of 100 most abundant X.
laevis proteins with 100 most abundant proteins of 47 human tissues and cell lines led to detection of two RNA binding proteins that are not significantly expressed at human tissues and cell lines (Wilhelm et al. 2014). These are cold-inducible RNA-binding protein B (CIRBP) and Y-box-binding protein 2 (YBX2). Both of them are essential for embryonic development. On one hand, YBX2 binds to mRNA and marks mRNAs for cytoplasmic storage (Tafuri & Wolffe, 1990). Highly abundant expression of YBX2 makes perfect biological sense for an organism to store enormous amounts of RNA. On the other hand, CIRBP may act negative regulator for adenylation of mRNAs, thereby preventing their translation (Aoki et al. 2003). This is an essential mechanism to tune translation capacity of stored mRNAs in X. laevis oocyte and embryos in the absence of transcription (Mendez & Richter 2001). Exportin 6 and c-Mos are prominent examples of poly(A)-tail-based translation regulation (Bohnsack et al. 2006; Gebauer et al. 1994). As a result, 100 most abundant X. laevis proteins are involved in common biological processes, such as protein translation, and energy production. However, presence of two highly abundant RNA binding proteins highlights that biological specificity of the oocytes can be reflected by presence of signature proteins.
Spatial proteome analysis revealed not only qualitative but also quantitative differences between the cytosolic and the nuclear proteomes. Quantitative localization distribution of the cytosolic and nuclear proteomes revealed tri-modal distribution. Two-third of the identified proteins were found in both compartments, unexpectedly only a small fraction (~500 proteins) of identified proteins had equimolar distribution in both compartments (Figure 4.10B). On the other hand, almost one-third of the identified proteins were unique to each compartment, where the nucleus contained both higher number of compartment specific and total protein species (Figure 4.10B).
The comparative analysis of 100 most abundant proteins in the cytosolic, the nuclear and the total oocyte fraction revealed dramatic proteome difference among those. Hierarchical clustering revealed four distinct proteins groups based on their abundance. Abundance based clustering indicates which molecular pathways are dominated in which compartment. The Cluster I composed of proteins, which are highly abundant in the cytosol, and to some extent (apart from Cluster I’) in the nucleus (Figure 4.11A). Majority of these proteins were ribosomal proteins, glycolysis pathway members and some chaperones. Since ribosomal biogenesis starts in the
group of proteins, which were highly abundant in all fractions (Figure 4.11A, Cluster I’). These proteins indicate cellular processes, which are predominantly takes place both in the cytosol and in the nucleus. One of them was actin, due to the lack of Exportin 6 expression till egg stage; actin can diffuse into the nucleus, and it forms filaments to provide mechanical stability to the giant nucleus (Bohnsack et al., 2006). GTP-binding protein Ran is another high abundant which were present in almost equal concentration in the cytosol and in the nucleus. Considering essential role of the Ran in nucleocytoplasmic transport equimolar distribution of is anticipitated. Another protein is ubiquitin; polyubiquitination is signal for protein degradation that I previuoslz shown that ubiquitin proteasome system is not spatially restricted (Figure 4.18). Interestingly, there were several glycolysis pathway members and ATP producing enzymes, such as creatine kinase, nucleoside diphosphate kinase A1, phosphoglycerate kinase. High nuclear abundance of several ATP producing enzymes raises the possibility of local ATP production in the nucleus. Due to the enormous size of the nucleus, passive diffusion of ATP would be inefficient to supply all energy requirement of the oocyte. Apart from these ATP producing enzymes, several glycolysis pathway members were highly abundant in the nucleus, such alpha-enolase and transketolase.
Nuclear proteome was dominant with compartment specific proteins involved in DNA replication, chromatin organization, and RNA processing. On the other hand, the cytosol contains predominantly proteins involved in protein synthesis, like translation initiation factors, and aminoacly tRNA synthetases. Apart from these, cytoskeleton and cytoskeleton-associated proteins, such as actin regulators and motor proteins were highly abundant. As a result, spatial proteome of X. laevis analysis underlines that compartmentation results in spatial distribution of distinct molecular activities.