a Whole Genome Array Approach - Jacobs University

Results and Discussion
striking feature of the expression profiles displayed is that a large fraction of the genome
responded in a stereotypical manner to all three stressful conditions tested. In summary 152
genes are up- or down-regulated at any time point for all stressors. Out of the 152 genes 62
are induced and 90 are repressed (TABLE 2 & TABLE 3). 49% of the induced and 61% of
the repressed genes were annotated as hypothetical proteins. The Venn diagrams shown in
FIGURE 3 provide an overview of the specific and common genes of the three stress-specific
responses. To identify co-regulated patterns of gene expression, we classified all
differentially expressed genes of all three stress expressions into 30 k-means clusters based
on their expression log ratio. The cluster data are available in the ADDITIONAL FILE 2.
Cluster 1, 3 and 4 show a similar response to the specific environmental changes, called
environmental stress response (ERS) over all experiments. Cluster 2, 4, 5, 7, 15 and 22
revealed genes that are reacting only to one or inverse to a specific environmental factor.
Experimental design and assessment of array data quality
The experimental conditions were chosen to mimic the natural environment of R. baltica with
the constraint that the stress response must be detectable by the microarray technology and
should therefore be significantly pronounced. In contrast to steady-state or single-time-point
studies, time series experiments can provide useful data for generating computational models
of stress response pathways [29].
The negative, positive and stringency controls printed on the array gave no indications for
unspecific hybridizations. Co-hybridizations of two cDNA samples prepared from the same
total cellular RNA (self-self hybridization) suggested that genes with an expression log ratio
value greater than 1.5 and smaller than -1.0 for heat and cold shock, respectively, could be
regarded as differentially expressed. For the salt stress log ratio values over 1.2 and
below -1.0 were considered as significant.
Effect of stress on Rhodopirellula baltica
No growth was detectable during stress conditions nor was any obvious morphology change
visible under the microscope. Under optimal conditions R. baltica has a doubling time of 10-
12 hours [22], indicating that physiological effects are not measurable during the short time
period of max. 5 hours the cells were exposed to stress.
28