​tu-bs.​de/​; [12]]. Many Roseobacter strains, including R. denitrificans, R. litoralis, Dinoroseobacter shibae and S. pomeroyi carry plasmids of different size [13, 14]. They range from 4.3 kb to 821.7 kb and can carry up to 20% of the genome content [4]. Therefore, due to Torin 1 mouse possible incompatibilities, the choice of suitable vectors for genetic investigations is of enormous importance [15]. The

availability of the complete genome sequences of this important group of bacteria is a crucial prerequisite for a detailed analysis of their physiological and ecological properties. However, for systems biology approaches suitable methods allowing easy and efficient genetic manipulation of these strains are needed. Such techniques are already established for other members of the Rhodobacteraceae, including Rhodobacter sphaeroides and Rhodobacter capsulatus [e.g. [16–18]]. However, in this context only little is known for members of the Roseobacter clade. Techniques for electroporation, transposon mutagenesis, biparental mating, gene knockout and genetic complementation were described only for Silicibacter sp. TM1040 [19, 20], S. pomeroyi [21, 22] and Sulfitobacter sp. J441 [23].

In the latter study, also lacZ reporter gene fusions were constructed for gene expression analyses. Moreover, transposon mutagenesis of Phaeobacter sp. was described [19]. However, already in 2005, the Roseobacter clade comprised a large phylogenetic diversity with 36 described species representing 17 genera [6]. In the meantime, many more species have been described, making it increasingly difficult CYC202 in vitro to obtain stable tree topologies based on 16S rRNA sequences [4]. It is well known from other bacterial groups that genetic tools developed for one genus do not work in a related genus or even in a different strain of the

same species. Therefore, we systematically determined key parameters required for successful genetic experiments in strains which cover phylogenetic groups Paclitaxel cost complementary to the few already studied. We selected R. litoralis and R. denitrificans, the archetypical isolates from the Roseobacter clade whose physiologies have been studied for a long time. Moreover, Oceanibulbus indolifex, a non phototroph which is related to Sulfitobacter was selected. All three species are in the middle of the Roseobacter radiation [4]. Furthermore, we selected two species of Phaeobacter (formerly Ruegeria). Finally, D. shibae a genus which is at the base of the Roseobacter radiation, was studied in more detail. We first investigated the antibiotic susceptibility of the selected Roseobacter clade species to identify useful selective markers. Using these antibiotic markers, we tested transformation and Compound C cost conjugation methods using plasmid-DNA transfer with different classes of plasmids.