Therefore, sliding means for 20 adjacent dots were calculated and plotted to help visualise patterns (red

dots, Figure 5). Again no general relationship between position along one axis and position along the other could be established. Nevertheless the ori and right loci appeared STI571 in vivo to behave similarly and the NS-right locus tended to be closer than ori and right to the cell centre. The ter locus was more SGC-CBP30 purchase peripheral than other loci in cells with a single focus (red dots). The same analysis was performed for the ori and ter loci after Ndd treatment (Figure 5). For the ter locus, distributions of the two cell classes were combined since they were not significantly different (Additional file 1, Figure S4D). In both cases, the sliding mean was consistent with the peripheral location of the loci. Equivalent patterns were obtained for the right and NS-right

loci in Ndd-treated cells (not shown). Foci located in the 0-0.1 cell length slice were more central than the other foci. This cell length slice corresponds ROCK inhibitor to the cell poles, where the membrane curvature modifies the cell width distribution of foci. This effect was detected only in Ndd-treated cells due to the enrichment of loci in this cell slice compared to control cells (Additional file 1, Figure S4C). Figure 5 Analysis of correlation of the position of foci along the cell length with that along the cell diameter. Graphs show the positions of foci of four loci in wt and Ndd-treated cells, as indicated in each panel, along the cell diameter (Y-axis) as a function of their position along the cell length (X-axis). The grey dots are individual foci. The red dots are sliding means of twenty adjacent foci (with a step of one focus). For the ori, right and NS-right loci in Ndd-untreated cells and for the ori and ter loci in Ndd-treated cells, the data from the different cell classes were combined, as these dataset do not statistically differ (see Figure 2). In the case of the ter locus in Ndd-untreated cells, only the data from cells with a single

focus are plotted. The dotted lines show the mean position of foci calculated from the 90% central model. Discussion We report that it is possible to assess the mean position of chromosome loci across the width of a rod-shaped bacterium using two-dimensional oxyclozanide pictures. We recorded the apparent position of fluorescence-tagged chromosomal loci along the diameter of a large number of cells and compared the resulting distributions to simulated distributions calculated from different positioning models. We analysed five loci mapping in four different chromosomal regions that behave differently during the cell cycle. For these five loci, we detected three different patterns, showing that our method can detect differences in cell width localisation. The ori and right loci appeared randomly distributed through a cell volume corresponding to the nucleoid, whereas the NS-right locus was more central and ter loci more peripheral.