Oxidative stress and starvation in Dinoroseobacter shibae: the role of extrachromosomal elements

Abstract

Aerobic anoxygenic phototrophic bacteria (AAP) are abundant in the photic zone of the marine environment. Dinoroseobacter shibae, a representative of the Roseobacter group, converts light into additional energy that enhances its survival especially under starvation. However, light exposure results in the production of cytotoxic reactive oxygen species in AAPs. Here we investigated the response of D. shibae to starvation and oxidative stress, focusing on the role of extrachromosomal elements (ECRs). D. shibae possessing five ECRs (three plasmids and two chromids) was starved for 4 weeks either in the dark or under light/dark cycles and the survival was monitored. Transcriptomics showed that on the chromosome genes with a role in oxidative stress response and photosynthesis were differentially expressed during the light period. Most extrachromosomal genes in contrast showed a general loss of transcriptional activity, especially in dark-starved cells. The observed decrease of gene expression was not due to plasmid loss, as all five ECRs were maintained in the cells. Interestingly, the genes on the 72-kb chromid were the least downregulated, and one region with genes of the oxygen stress response and a light-dependent protochlorophyllide reductase of cyanobacterial origin was strongly activated under the light/dark cycle. A Δ72-kb curing mutant lost the ability to survive under starvation in a light/dark cycle demonstrating the essential role of this chromid for adaptation to starvation and oxidative stress. Our data moreover suggest that the other four ECRs of D. shibae have no vital function under the investigated conditions and therefore were transcriptionally silenced.

Keywords: aerobic anoxygenic photosynthesis; chlorophyll a biosynthesis; light-stress adaptation; plasmid-maintenance; reactive oxygen species; starvation.

Figure 1

Experimental design of the starvation experiment. (A) Total cell counts of D. shibae during 4 weeks of starvation and the time points used for microarray analysis. Gray squares represent the light/dark (LD) while the black squares represent the continuous dark (DD) starved samples. The continuous line represents the total cell counts from the current study. Data from our previous study (Soora and Cypionka, 2013) are shown in the dashed lines. The black arrows indicate the sampling points. (B) Microarray loop design. Boxes represent samples and are labeled according to their starvation day and condition. The dark and light gray boxes represent the conditions DD and LD respectively. Two color microarrays are represented by arrows connecting the Cy3 and Cy5 labeled samples that are being compared. The continuous arrows mark the microarrays analyzed further.

Figure 2

Substantial changes in gene expression of D. shibae. Genes with significant changes in expression on the chromosome and on extrachromosomal replicons (ECRs) of D. shibae when (A) LD vs. DD, (B) LD vs. NS, and (C) DD vs. NS are compared. Genes that are part of the photosynthesis gene cluster and the oxidative stress response are highlighted in green and red, respectively.

Figure 3

Overview of differential gene expression of D. shibae cells starved for two days. (A) log₂-fold changes in chromosomal genes; the upper box represents the light/dark cycle (LD) and the lower portray the dark (DD) starved cells. (B) The theoretical quantile-quantile (q–q) plot is shown for both LD (gray) and DD (black) chromosome genes, which are generally used to compare the distribution, showed that they are normally distributed. (C) log₂-fold changes in the ECRs. The upper lane represents the q-q plot and the lower represents the gene expression pattern of the ECRs. Expression under light/dark cycling (LD) and continuous dark (DD) is shown for each gene with gray and black circles, respectively.

Figure 4

Expression of the regulated locus on the 72-kb chromid. The upper bar chart illustrates the differential expression of the regulated locus on the 72-kb chromid (Dshi_4160-4172). In the lower panel, the light dependent chlorophyllide reductase (LPOR) and genes that have been found to be part of the oxidative stress response (Tomasch et al., 2011) are highlighted in green and red colors respectively.

Figure 5

PCR assay for the detection of the five extrachromosomal elements and the RNA expression profile. (A) Agarose gels with PCR amplificate of 12 randomly selected colonies of 4 weeks starved D. shibae cells. The sizes of the PCR products for ECRs are 201 bp (191-kb), 205 bp (153-kb), 201 bp (126-kb), 208 bp (86-kb) and 201 bp (72-kb). E. coli DNA was used as a template for negative control (NC). (B) Gene transcript profiles for the five ECRs based on qRT-PCR.

Figure 6

Microscopic and graphical representation of viable D. shibae wild-type and ECR cured cells. (A) Live/Dead staining where green and red color denotes viable and dead cells, respectively. (B) Viability of wild type and plasmid cured mutants (Δ72-kb and Δ86-kb) under LD and DD condition during 3 weeks of starvation.

Figure 7

Light-dependent chlorophyllide reductase (LPOR) in marine Roseobacter and Erythrobacter strains. (A) Maximum likelihood tree of the LPOR protein sequences of marine Cyanobacteria, Roseobacter and Erythrobacter strains. The latter ones are highlighted in red. 1000 bootstrap replicates were performed. The percentage of trees in which the associated taxa clustered together is shown next to the branches. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. (B) Genomic neighborhood of the LPOR gene in Roseobacter strains and Erythrobacter litoralis. Genes discussed in the manuscript are highlighted.