Decoding the nucleoid organisation of Bacillus subtilis and Escherichia coli through gene expression data

Abstract

Background: Although the organisation of the bacterial chromosome is an area of active research, little is known yet on that subject. The difficulty lies in the fact that the system is dynamic and difficult to observe directly. The advent of massive hybridisation techniques opens the way to further studies of the chromosomal structure because the genes that are co-expressed, as identified by microarray experiments, probably share some spatial relationship. The use of several independent sets of gene expression data should make it possible to obtain an exhaustive view of the genes co-expression and thus a more accurate image of the structure of the chromosome.

Results: For both Bacillus subtilis and Escherichia coli the co-expression of genes varies as a function of the distance between the genes along the chromosome. The long-range correlations are surprising: the changes in the level of expression of any gene are correlated (positively or negatively) to the changes in the expression level of other genes located at well-defined long-range distances. This property is true for all the genes, regardless of their localisation on the chromosome. We also found short-range correlations, which suggest that the location of these co-expressed genes corresponds to DNA turns on the nucleoid surface (14-16 genes).

Conclusion: The long-range correlations do not correspond to the domains so far identified in the nucleoid. We explain our results by a model of the nucleoid solenoid structure based on two types of spirals (short and long). The long spirals are uncoiled expressed DNA while the short ones correspond to coiled unexpressed DNA.

Figure 1

Illustration of the methodology used in this study. Example of the results obtained on a hypothetical bacterial circular chromosome model of 4 genes. The gene expression intensities are measured in three experimental conditions. Part 1 is normalised data (mean equal to 0 variance equal to 1) according to experimental conditions. Part 2 is the matrix of Kendall tau (see methods). Part 3 is the autocorrelation matrix with inter-gene distances. Part 4 is the averaged linear autocorrelation.

Figure 2

Long-range averaged autocorrelations in B. subtilis. To identify the regularities which are common to most of the genome, regardless of the genes localisation, the autocorrelation vectors of all the genes were summed (blue curve). This global signal shows the averaged autocorrelation regularities as a function of inter-gene distance. The green curve shows the averaged autocorrelation when the genes positions on the genome were randomly assigned. The red curve represents the resultant of four oscillations of periods 600 ± 55, 240 ± 21, 113 ± 21 and 60 ± 6 genes, which were estimated from the averaged autocorrelation deconvolution. The horizontal scale represents the distance between two genes (the difference of their ranks on the chromosome). The green, blue and red curves have the same vertical scale. The red curve was shifted for readability. Whereas the green signal shows no regularity, long-range correlations can be seen in the blue signal (maxima at ca. 200, 650, 850, 1300, 1500 and 2050 inter gene distance and minima at ca. 550, 900 and 1750–1950).

Figure 3

Short-range co-expression regularities in B. subtilis. To identify the regularities which are common to most of the genome, regardless of the genes localisation, the autocorrelation vectors of all the genes were summed (blue curve). This global signal shows the averaged autocorrelation regularities as a function of inter-gene distance. The green zone shows the averaged autocorrelation when the genes positions on the genome were randomly assigned (mean of the random signal ± the root mean square deviation). The horizontal scale represents the distance between two genes (the difference of their ranks on the chromosome). Neighbouring genes on the chromosome show highly correlated variations of expression levels. The averaged autocorrelation of two contiguous genes is 0.4. The signal can be decomposed into two parts: (i) inter-gene distances between 1 and 5–6 genes are characterised by a high autocorrelation, which drops steeply; (ii) beyond 6 genes the autocorrelation shows a regular and slower decrease. The autocorrelation merges with the background noise at an inter-gene distance of about 100 genes (similar to 100 kb). The autocorrelation decrease may be seen as the resultant of a linear decrease and 14.5 ± 1 genes period oscillations (red curve).

Figure 4

Partial sums of the autocorrelations in B. subtilis. To analyse if the discovered regularities depend on gene position, the autocorrelation vectors of groups of 500 genes were summed up (9 coloured curves). The horizontal scale represents the distance between two genes (the difference of their ranks on the chromosome). All the curves were vertically shifted for readability. The signals show the co-expression regularities according to inter-gene distance. Long-range periodicities are shared by all the signals regardless of the gene groups.

Figure 5

Long-range averaged autocorrelation in E. coli. To identify the regularities which are common to most of the genome, regardless of the genes localisation, the autocorrelation vectors of all the genes were summed (blue curve). This global signal shows the averaged autocorrelation regularities as a function of inter-gene distance. The green curve shows the averaged autocorrelation when the genes positions on the genome were randomly assigned. The red curve represents the resultant of two oscillations of periods 557 ± 30 and 100 ± 18 genes, which were estimated from the averaged autocorrelation deconvolution. The horizontal scale represents the distance between two genes (the difference of their ranks on the chromosome). The green and blue curves have the same vertical scale. The red one is on a scale, which is moved down for readability. Whereas the green signal shows no regularity, long-range periodicities can be seen in the blue signal (maxima at ca. 200, 650, 1100, 1400 and 1700 and minima at ca. 850, 1380 and 2180).

Figure 6

The possible chromosome configuration. We assume that the nucleoid structure consists of a solenoid with two types of spirals: • Large spirals of uncoiled DNA, containing the genes that are transcribed, that lie on the surface of the nucleoid and define its diameter (0.5 μm). • Small spirals of coiled untranscribed DNA that lie inside the nucleoid.