The presence of five different plasmids in Sphingomonas sp. MM-1 clearly demonstrated that there must

exists at check details least five different incompatibility groups in sphingomonads, and it can be assumed that the pronounced rearrangements, which occur after the conjugative transfer of degradative plasmids among sphingomonads, might be (at least in certain cases) related to incompatibility phenomena (Feng et al., 1997a, b; Ogram et al., 2000; Basta et al., 2004, 2005). The phenotypically defined incompatibility groups can be correlated with the sequences of the replication initiator (Rep) proteins and the proteins involved in plasmid partition (Par) (Petersen, 2011). In this context, the Rep proteins are especially important as these are responsible for the initial site specific DNA-binding and nicking activities, which represent the first steps in plasmid replication. The plasmid sequences deposited at the NCBI database originating from the genera Sphingomonas, Sphingobium, Novosphingobium and Sphingopyxis clearly demonstrated that the genes annotated as rep genes (repA or repB) almost exclusively belong to three protein superfamilies. Thus, proteins belonging to the RepA_C superfamily (Pfam 04796), Rep_3 (Pfam 01051) and

RPA superfamily (Pfam 10134) were found in large numbers among the deposited sequences (Table 1). In addition, the Rep proteins from four smaller plasmids (pUT2, pYAN-1, pSx-Qyy, Spl) Talazoparib in vivo – which do not carry any catabolic genes – were classified to belong to the HTH-36 superfamily (Pfam 13730). An alignment of the Rep-sequences allowed the construction of a dendrogram to visualize the relationship among the different Rep-sequences (Fig. 1). This demonstrated that the Rep proteins from the large degradative plasmids

pNL1, pCAR3, pSWIT02 and Mpl can be clearly differentiated from the Rep proteins encoded by other plasmids. Thus, these Rep proteins belong to SPTLC1 the RepA_C family and are composed of about 430 amino acids (aa). In contrast, all other annotated Rep proteins are almost consistently smaller than 400 aa and did not belong to the RepA_C superfamily (Table 1). A second group of ‘megaplasmids’ consists of plasmid pISP1 (172 kbp) from Sphingomonas sp.MM-1, pNL2 (487 kbp) from N. aromaticivorans F199 and Lpl (192 kbp) from Novosphingobium sp. strain PP1Y. These plasmids encode for Rep proteins belonging to the RPA superfamily. Obviously, the plasmids of this group (=‘Mega-RPA’) are compatible with those of the group defined above (=‘Mega-RepAC’) as plasmids pNL1 and pNL2 are found together in N. aromaticivorans F199, and plasmids Mpl and Lpl in Novosphingobium sp. strain PP1Y (Romine et al., 1999; D’Argenio et al., 2011). A third group of large degradative plasmids was identified among the plasmids that possess a Rep protein belonging to the Rep_3 superfamily.