Transcriptome analysis revealed that a quorum sensing system regulates the transfer of the pAt megaplasmid in Agrobacterium tumefaciens

Abstract

Background: Agrobacterium tumefaciens strain P4 is atypical, as the strain is not pathogenic and produces a for this species unusual quorum sensing signal, identified as N-(3-hydroxy-octanoyl)-homoserine lactone (3OH,C8-HSL).

Results: By sequence analysis and cloning, a functional luxI-like gene, named cinI, has been identified on the At plasmid of A. tumefaciens strain P4. Insertion mutagenesis in the cinI gene and transcriptome analyses permitted the identification of 32 cinI-regulated genes in this strain, most of them encoding proteins responsible for the conjugative transfer of pAtP4. Among these genes were the avhB genes that encode a type 4 secretion system (T4SS) involved in the formation of the conjugation apparatus, the tra genes that encode the DNA transfer and replication (Dtr) machinery and cinI and two luxR orthologs. These last two genes, cinR and cinX, exhibit an unusual organization, with the cinI gene surrounded by the two luxR orthologs. Conjugation experiments confirmed that the conjugative transfer of pAtP4 is regulated by 3OH,C8-HSL. Root colonization experiments indicated that the quorum sensing regulation of the conjugation of the pAtP4 does not confer a gain or a loss of fitness to the bacterial host in the tomato plant rhizosphere.

Conclusion: This work is the first identification of the occurrence of a quorum sensing regulation of the pAt conjugation phenomenon in Agrobacterium.

Keywords: Acylhomoserine lactone; Agrobacterium; At plasmid; Conjugation; Dtr system; Quorum sensing; T4SS; Transcriptomics.

Fig. 1

Identification of N-(3-hydroxy-octanoyl)-homoserine lactone (3OH,C8-HSL) in culture supernatants of strain P4. An ethyl acetate extract of a spend culture supernatant was concentrated, resuspended in 100 % acetonitrile and analyzed by UPLC MS/MS. Panel a: elution profile of synthetic 3OH,C8-HSL. Panel b: mass spectra of synthetic 3OH,C8-HSL. The m/z values 244, 226 and 102 correspond to the protonated forms of 3OH,C8-HSL (243 + 1), the dehydrated 3OH,C8-HSL (225 + 1) and the characteristic amine substituted lactone cycle (101 + 1), respectively. Panel c: elution profile of the concentrated culture supernatant of strain P4. The major peak migrates as the one of synthetic 3OH,C8-HSL. Panel d: mass spectrum of the major peak observed in the concentrated culture supernatant of strain P4. The m/z values and species are identical to those observed for synthetic 3OH,C8-HSL

Fig. 2

Genetic determinants of the quorum sensing regulation on pAtP4 and related regions of other plasmids. Panel a: genetic organization of the cinI region on the pAt plasmid of A. tumefaciens strains P4 and 5A, A. radiobacter DSM 30147 and in the genome of R. leguminosarum strain 3841. Gene sizes are drawn to scale (scale at the left bottom part of panel a). Panel b: Phylogenetic relatedness of the deduced CinI protein of strain P4. The tree has been constructed using the Neighbor-Joining method and bootstrap values calculated from 1,000 re-samplings technique. Panel c: Phylogenetic relatedness of the deduced CinR and CinX proteins of strain P4. The tree has been constructed as indicated above

Fig. 3

Comparison of three pAtP4 regions displaying upregulated genes. Fold change of pAtP4 genes obtained from the transcriptome experiments are shown above each panel. Upregulated genes of the pAtP4 are shown by arrows with different backgrounds while the non-regulated genes are shown by open arrows. Gene sizes are drawn to scale (scale at the left bottom part of panel c). Panel a: the avhB cluster of pAtP4 with genes upregulated or not. The avhB cluster of pAtP4 was aligned with the avhB cluster of pAtC58 and compared with the trb cluster of pTiC58. C58 genes displaying significant similarity to avhB genes are marked with the same background pattern as used for the corresponding avhB gene of pAtP4. Panel b: the tra cluster of pAtP4 with genes upregulated or not. The tra cluster of pAtP4 was aligned with tra cluster of pAtC58. Up regulated and non-regulated genes are represented as indicated above. Panel c: the cinR cinI cinX genomic region with upregulated genes of the pAtP4. The cinR cinI cinX genomic region was aligned with the corresponding pAt region of A. radiobacter strain DSM 30147 and compared to the one of R. leguminosarum bv. viciae strain 3841. Up regulated and non-regulated genes are represented as indicated above

Fig. 4

pAt conjugation efficiencies of strain P4 and related mutants. Crosses involved the pAt donor indicated on the x axis and strain C58.00 as a recipient. Conjugation frequencies (y axis) were expressed per recipient cells. A Kruskal-Wallis non parametric test and a post-hoc Tukey test were used to evaluate whether the observed differences were significant (p ≤ 0.05) or not (p > 0.05). Results are shown by the letters displayed above the graph

Fig. 5

Fitness of P4 and the P4 cinR cinI cinX mutant in the tomato rhizosphere. Cell densities (CFU/g of soil) of strains P4RifNC and P4RifRIX inoculated simultaneously on tomato plants cultivated in an unsterilized horticultural soil were assessed by dilution and plating onto the appropriate media (see Methods). Panel a: cell density of both co-inoculated P4RifNC and P4RifRIX. Panel b: ratio of the cell densities of each strain. At each time point (but not between time points), the values generated by the 3 repeats were statistically analyzed by a Mann and Whitney non parametric test which indicated that the proportion of the cell densities of the two strains were not statistically different (p-values > 0.05)