2g). To conclude, it is apparent that GFP-MinDEc is able, at least partially, to substitute the role of MinDBs during B. subtilis cell division. As a positive control, we inspected ΔminDBs strain expressing GFP-MinDBs (IB1059) in a similar way as described above for GFP-MinDEc. Without addition of xylose, GFP-MinDBs was able to improve the phenotype of ΔminDBs cells (Fig. 2h) and the average cell length decreased to 3.3 μm. In addition to cell morphology, the localization

of GFP-MinDEc in a wild-type background (IB1103), in ΔminDBs (IB1104) and in ΔminDΔdivIVA (IB1105) cells was examined by fluorescent microscopy. We noticed a high level of background fluorescence in the cytosol, indicating a possible GFP-MinDEc fusion proteolysis. This was confirmed using Western blot analysis (Fig. 3a). HDAC inhibitor The background fluorescence signal was not prevented when the cells were grown at a lower temperature (28 °C) (data not shown). A strain with YFP-MinDEc fusion, expressed from Phyperspank promoter, was prepared to

improve the localization images. This gene fusion was introduced into the amyE locus of MO1099, creating the strain IB1110; into IB1056 (minDBs::cat) and IB1109 (minDBs::cat divIVA::tet) generating IB1111 and IB1112 strains, respectively. The resolution was clearly improved and the fluorescence background level was decreased, indicating that the YFP-MinDEc fusion protein was more stable than GFP-MinDEc, as confirmed by Western blot analysis (Fig. 3b). Moreover, the expression from this promoter seems to be controlled SRT1720 cost more tightly than from Pxyl promoter because no signal was visible in the absence of IPTG when examined by Western blot analysis (Fig. 3a and b). Under the lowest expression level tested (0.1 mM IPTG) the average cell length of the strain IB1111 (minD::cat, amy::Phyperspank-yfp-minDEc) decreased to 3.2 μm. This is a better complementation result than observed for strain IB1104 (minD::cat, amy::Pxyl-gfp-minDEc). In all three strains (IB1110, IB1111 and IB1112) the observed YFP-MinDEc signal suggested the existence

of helices winding along the cell length. However, in some cells the signal was present as dots at the membrane, or at cell poles and potential division sites (Fig. 4a). The strains were also examined for the potential dynamic behaviour of the YFP-MinDEc using time-lapse CYTH4 microscopy. The images were taken every 10 s for 2 min. It was not possible to observe the oscillatory movement of either GFP-MinDEc or YFP-MinDEc. To find out whether YFP-MinDEc can recognize the same membrane system as GFP-MinDBs in B. subtilis, the cells were stained with FM 4-64, which preferentially stains negatively charged phospholipids (Barák et al., 2008). In the overlay picture the green (representing YFP-MinDEc) and red (representing FM 4-64) fluorescence signals, which are in close proximity, become yellow (Fig. 4b). Most of the YFP-MinDEc signals clearly colocalize with FM 4-64 fluorescence.