Abstract

Under starvation conditions, the soil bacterium Bacillus subtilis undergoes a complex program geared towards survival, which ultimately results in the formation of highly resistant yet dormant endospores. These spores contain packed DNA encased by a peptidoglycan cortex and concentric proteinaceous layers, protecting the spore from environmental influences. The crust is of special interest in this thesis, as it constitutes the outermost layer of the coat. In this thesis, the goal is to display proteins of interest onto the surface of the spore, thereby creating biologically active particles, namely Sporobeads. For this purpose, it is only logical to utilize the spore’s outermost layer: the crust. Moreover, this thesis aims to not only provide a proof of concept, but to also establish an easy-access vector system (Sporovectors) to easily produce Sporobeads for any given application. As scientific knowledge on the nature of the crust still has major gaps, this thesis aims to also reach a deeper understanding of this particular layer, which in turn might help to better understand and improve the application of the Sporobeads. Therefore, the interaction network of the six crust proteins (CotVWXYZ and CgeA), their role in the crust as well as the composition and the active players of the proposed glycosylation modification were also elucidated in this study. These two projects benefit from each other in a create-test-learn-redesign cycle. With the established methodology of the Sporobeads in place and the first insights into the crust proteins at hand, the basic research into the crust is more convenient. In turn, the nature of the crust helps explain the performance of the anchors and also provides a rationale for further improvement strategies. To this end, the study showed that the utilization of the crust as a platform for protein display is feasible. The best anchors were CotY and CotZ, followed by CotX and CotV. CotW and CgeA were the least promising. The enzymes were stabilized on the spore surface during storage, and it was also possible to recycle the particles. The relative performance of the different anchors is partially explained by the protein interaction network and their roles in the crust structure. CotY and CotZ are the major structural pillar, whereas CotX and CotV play a more supportive role in structure propagation. CotW supports the assembly of the CotX/CotV structural pillar. CgeA seems to play a role in glycosylation, being least abundant. CotZ anchors the crust structure to the middle part of the spore, whereas CotX and CotY might already anchor it loosely to the poles of the spore. CotX and CotV appear to be the most probable candidates for glycosylation, due to the conserved glycosylation motif in the CotX superfamily domain. The mode of glycosylation is quite complex, involving many players and presumably at least six different sugars divided into two independent yet probably cross-linked polysaccharide species: one related to rhamnose and one to galactose. The putative functions of the players involved seem to indicate that the rhamnose-related species might contain the rare sugar viosamine or VioNAc. Furthermore, the rhamnose-related polysaccharide variant could also potentially be capped by a unique sugar based on viosamine or VioNAc modified with a lysine side chain, similar to what happens in a relative Bacillus anthracis. This suggests that water dispersal is not the only role of this modification, but that said modification might act as additional protection against biological scavenging, due to its exceptional nature. Lessons learned regarding the nature of the crust can, in turn, foster the development of potential improvement strategies. On the whole, however, these strategies turned out to be less promising than was initially hoped. Linkers were able to rescue variants performing inadequately, but only led to slight improvements for the best-working variants. On the other hand, removing native competition only slightly improves the performance. This is probably due to the high level of redundancies (CotY/CotZ and CotV/CotX) as well as the high level of interdependencies in the crust. When considering that the crust is glycosylated, some enzymes might be less active or even inactive in this micro-environment. With this rationale, changing the surface properties using a mutant with an impaired crust polysaccharide structure (cgeA) slightly improved the activity of the Bacillus pumilus laccase. This might, in future, enable the performance of some enzymes requiring a more hydrophobic environment, such as lipases. Nonetheless, a peculiarity discovered during this study could lead to a novel field of application: spore-derived self-assembled non-GMO including particles (SporoSNIPs). In some mutants, the assembly of the crust proteins onto the spore was highly disturbed, but the crust proteins themselves self-assembled in the mother cell instead (in the cotZ mutant the complete crust structure, or CotX and CotV in any mutant lacking CotY or CotZ). These fragments (SporoSNIPs) could potentially be enriched or completely separated from the spores for further utilization as biologically active particles, but without the disadvantage of containing living GMOs.