Moreover, the ORF 28 is homologous

to the ptmG gene of Campylobacter jejuni (Cj1324) which converts the CMP-Leg5Ac7Ac residue to CMP-5-acetamidino-7-acetamido-3,5,7,9-tetradeoxy-D-glycero-D-talo-nonulosonic acid (CMP-Leg5Am7Ac) [40], the dominant residue of the O-antigen of non-Sg1 strains of L. pneumophila[41]. A functional correlation of the ORFs of this region is supported by recent transcriptomic data of strain Paris in which the ORFs 21-17 and 28-22 were transcribed as operons [42]. Since all analyzed Sg1 strains and a broad number of non-Sg1 strains carry ORF 28 [35, 43, 44] it can be assumed that CMP-Leg5Am7Ac is a common residue of the L. pneumophila LPS this website molecule which might subsequently become modified in a mAb-subgroup or even strain specific selleck chemical manner. Three clusters of the O-acetyltransferase Lag-1 A well examined phenotype variation is linked to the presence and absence of the lag-1 gene. Lag-1 encodes for an O-acetyltransferase that conferred reactivity with mAb 3/1 and is exclusively found in Sg1 strains. Our results revealed three clusters of the lag-1 genes, although without any detectable relation to the mAb-subgroup switch which supports recent findings [45]

(Figure  2A). Lag-1 was previously reported to be involved in mAb-subgroup switches of different strains. However, this was generally due to gene deletion or loss-of-function LXH254 in vivo mutations of lag-1[46–49]. Complete and functional lag-1 genes were present in all mAb 3/1+ strains and were absent in all mAb 3/1- next strains. Besides that, the Philadelphia subgroup strains (Philadelphia 1 and Paris) as well as the Knoxville-subgroup strain Uppsala 3 carried a transposase and a partial duplication of ORF

2 adjacent to lag-1. Bernander et al. reported the region from ORF 2 to ORF 3 as unstable [46]. Looping out of the intermediate located lag-1 gene is assumed to be a potential consequence. Under in vitro conditions the deletion of the lag-1 gene occurred at with frequency of 10-6 to 10-7 (C. Lück, unpublished results). Detailed analysis of the region from ORF 2 to ORF 3 including lag-1 of these strains revealed remarkably high similarities of Uppsala 3 to the Philadelphia-subgroup strains Philadelphia 1 and Paris (>98-100%) whereas the remaining Knoxville-subgroup strains clustered in a different group (Table  3; Figure  2A). The high similarity of this 4 kb region between strain Uppsala 3 and the strains Paris and Philadelphia 1 may indicate horizontal gene transfer of this region. However, this had no impact on the specific mAb reactivity for all other analyzed Knoxville-subgroup strains. Horizontal gene transfer between strain Paris and Philadelphia 1 was recently reported for a large genome fragment which also harbored the LPS biosynthesis locus [32].