The work presented in this thesis provides novel insights in several aspects of the molecular biology of archaea, bacteria and their viruses. Three fundamentally different groups of viruses are associated with the three domains of life. Archaeal viruses are characterized by a particularly high morphological and genetic diversity. Some archaeal viruses, such as Sulfolobus islandicus rod-shaped virus 2 (SIRV2), have quite remarkable infection cycles. As described in Chapter 1, infection with SIRV2 results in the formation of large virus associated pyramids (VAPs) on the host cell surface. The structures open in the final step of the infection cycle, creating large apertures to release the rod-shaped viruses that have matured in the cytoplasm. This virus release mechanism is unique and does not resemble egress mechanisms of bacterial and eukaryotic viruses. Analysis of the protein composition of SIRV2 infected cells, as outlined in Chapter 2, revealed the strong accumulation of the virus encoded protein PVAP in membranes after infection, suggesting involvement in VAP formation. The VAPs can be isolated as discrete particles, as demonstrated in Chapter 3. Electron microscopic survey of these particles showed that they are baseless pyramids with a heptagonal perimeter. This geometry is exceptional and especially the sevenfold symmetry is very rare in nature (20S proteasome, myosin). The structures can have various sizes, probably reflecting different developmental stages. This suggests that they grow by the gradual expansion of the triangular facets. Analysis of the protein composition of the structures revealed the exclusive presence of PVAP and anti-bodies raised against this protein labeled specifically the VAPs on thin sections of infected cells as observed in electron microscopy. PVAP is sufficient for VAP formation, which was demonstrated by expression of the protein and successful assembly of pyramidal structures, in the archaeon S. acidocaldarius and the bacteria Escherichia coli. Further analysis of PVAP truncation mutants as outlined in Chapter 4, showed that besides the 10 C-terminal amino acids, all domains of the protein are essential for VAP formation. PVAP can form oligomers of several sizes, including those of a heptamer, which probably act as nucleation points for VAP formation on the cell membrane. Analysis of the truncation mutants indicated that both the C and N terminal domain are important for interaction between monomers. Detailed observation with whole cell cryo-electron tomography of VAPs formed in the natural and heterologous system, revealed the presence of two layers in the structure. The outer one is continuous with the cell membrane. The inner layer facing the cytoplasm, presumably represents a protein sheet formed by tight interactions between the C-terminal domain of PVAP connected with a short linker region to the membrane. The sheets are slightly bended, giving the complete structure the appearance of a teepee. At the junction of two triangular sheets, the structure is perforated, creating predetermined breaking points. Furthermore, in this chapter data is presented which underlines the unique nature of this protein, since it is able to form VAPs successfully in archaeal, bacterial and eukaryotic membranes, which all fundamentally differ in protein and lipid composition. In case of expression in Saccharomyces cerevisiae, VAPs are formed on all membranes, including those of mitochondria, suggesting that the protein inserts spontaneously in membranes. Thus, PVAP serves as a universal membrane remodeling system, which might be exploited for biotechnological purposes, such as the development into a universal system for the controlled opening of ~100 nm apertures in any lipid bilayer. Production of VAPs is one of the dramatic consequences that SIRV2 infection has on the host cell. Whole transcriptome sequencing allowed determination of a global map of virus and host gene expression during the infection cycle, which is presented in Chapter 5. Directly after infection, transcription of viral genes starts simultaneously from both genome termini. All possible protein interactions between all SIRV2 proteins were assayed with yeast two-hybrid and these results were used to advance current knowledge on SIRV2 genes functions, of which the majority is still unidentified. The host cells respond to viral infection by adapting expression of more than 30% of its genes. Genes involved in cell division are down regulated, while those playing a role in anti-viral defense are activated. Specifically, for the first time massive activation of toxin anti-toxin and CRISPR-Cas systems is observed in an archaeal system. The different degree of expression and activation of the various systems highlights the specialized functions they perform. The CRISPR-associated multi-subunit ribonucleoprotein complexes that are crucial for the CRISPR mediated anti-viral defense, generally have an uneven stoichiometry, i.e. the 4-6 different protein subunits are present in different quantities. Just as most functionally related bacterial and archaeal genes, the cas genes are clustered in operons, which allow for co-expression (as has indeed been observed in the transcriptome analysis described in Chapter 5). This is advantageous when equal amounts of gene products are required, such as is the case for protein complexes with even stoichiometry. However, a substantial number of important protein complexes contain uneven stoichiometry. Employing comparative genomics, in Chapter 6, it is shown that differential translation is a key determinant of modulated expression of genes clustered in operons and that codon bias generally is the best in silico indicator of unequal protein production. In addition, analysis of protein production from genes with synonymous mutations from synthetic operons, provides evidence that initiation of translation can occur at intercistronic sites. The widespread occurrence of modulation of translation efficiency, suggests that this is a universal mode of control in bacteria and archaea that allows for differential production of operon-encoded proteins.