Molecular insights on the DNA-binding activity by the Escherichia coli Min CD proteins

Every living organism, including viruses, has to undergo reproduction in order to successfully establish a niche on this planet. One of the vital steps in the reproduction process is replication of the genetic material and its subsequent segregation between the two daughter cells. Errors in these processes may be fatal to the organism causing it to disappear from this planet. Eukaryotic cells have evolved an energy-consuming dedicated machinery to properly segregate the replicated genetic material between the two daughter cells. Unlike eukaryotes, prokaryotes adopt different mechanisms for chromosome segregation, which are in some cases still poorly understood. Moreover, in prokaryotes, chromosome replication and segregation happen simultaneously. While a dedicated chromosome segregation machinery has been identified in Caulobacter crescentus, such machinery has not been identified in many other prokaryotes. There is an ongoing debate in the prokaryotic field whether mere entropic forces of repulsion between the duplicated chromosomes alone can achieve full and precise chromosome segregation, or whether additional machineries are needed. A previous study from our group has shown that Escherichia coli MinD binds to DNA in a non-sequence specific manner. Based on computational, in vitro and in vivo analyses, it was hypothesized that such binding would be used by E. coli cells to properly segregate their chromosomes. Residues on MinD that, either directly or indirectly, affect DNA binding have been identified; however, the direct MinD-DNA binding interface is so far unknown. E. coli MinD, together with MinC and MinE, constitutes the so-called Min system, which has been extensively studied for its role in mid-cell determination. MinC actively counteracts FtsZ polymerization. It would do this anywhere in the cell, if it were not for MinD and MinE, which regulate MinC localization so that it is minimal at mid-cell, where the FtsZ ring can be assembled. The way in which MinD and MinE keep MinC away from mid-cell is very dynamic, and consists in pole-to pole oscillations that never cease. These oscillations are self-organized and occur also in the absence of MinC, as far as MinD, MinE and the membrane are present. Since MinC forms a complex with MinD, it effectively gets carried along in the oscillations, spending on average more time at the poles and being at low concentration at the centre. Without MinC, the Min oscillations do not exert any activity towards FtsZ. Thus, the three Min proteins ought to work together to achieve their goal of mid-cell determination. It is plausible that a concerted action of all Min proteins is required also for proper chromosome segregation, since we discovered that MinC strongly enhances the DNA-binding activity of MinD, while MinE terminates the binding by releasing MinC and the DNA off MinD (1). In my thesis, I aimed to identify the residues of MinC that either directly or indirectly aid MinD in DNA binding. To this end, potential MinC residues that could bind to the DNA were first computationally predicted and then mutated to experimentally test the consequences of such mutations on the DNA binding activity of MinC and MinD. Using electrophoretic mobility shift assay (EMSA) experiments, I found that glycine at position 10 and lysine at position 66 on MinC are involved in DNA binding since the MinCG10D and MinCK66A mutants showed strongly reduced binding. By performing circular dichroism experiments, I could exclude that the impairment of MinCG10D-MinD in DNA binding is due to changes in the secondary structure of the mutant protein, suggesting that the DNA is repelled by the negative charge of the aspartic acid. Most importantly, I discovered that not only the length, but also the amino acid sequence of the unstructured linker region of MinC, which connects the N- and the C-terminal domains of the protein, play a vital role in the DNA-binding activity of MinCD. Since MinC is also involved in inhibiting FtsZ polymerization, by performing cell viability spot assays I found that the linker should consist of at least two amino acids in order to efficiently inhibit FtsZ polymerization. EMSA assays with various synthetic constructs I made to test the necessity of different elements (MinC N- and C-terminal domain, linker region, MinD) for DNA binding revealed that the linker region of MinC is necessary for DNA binding. Further experiments are needed to understand if MinC N- and C-terminal domains and MinD are needed solely to place the linker in the proper orientation for it to bind to the DNA or if they contribute directly to the binding with specific residues. Interestingly, microscopy experiments performed using a synthetic construct made of the N-terminal MinC domain, the linker, a bZIP dimerizing domain and mRuby showed that this construct co-localizes with the E. coli nucleoid. Introduction of the G10D mutation on the synthetic construct did not alter its in vivo association with the nucleoid, suggesting that G10 is not used to directly bind to the DNA. Beyond studying the DNA binding activity of MinC and MinD, during my Ph.D. I analyzed the mechanism by which MinE is impaired when eYFP is C-terminally fused to it. By combining in vivo and in vitro assays, I show that eYFP makes the fusion protein prone to aggregation, and reduces the accessibility of MinE MTS as well as of arginine at position 21, needed to activate the ATPase activity of MinD. Finally, as to study biological processes it is often necessary to co-transform two plasmids in the cells of interest, I wanted to devise a method to reduce the requirement from two to one antibiotic to maintain two plasmids. To this aim, I employed split inteins to reconstitute full-length, functional enzymes conferring resistance towards antibiotics, which are expressed as two dysfunctional halves each on one plasmid. This method, which we called SiMPl, allows maintaining two plasmids in bacteria and mammalian cells using a single antibiotic chosen between kanamycin, chloramphenicol, ampicillin, hygromycin and puromycin