Research ArticleEVOLUTIONARY BIOLOGY

Repeated gain and loss of a single gene modulates the evolution of vascular plant pathogen lifestyles

See allHide authors and affiliations

Science Advances  13 Nov 2020:
Vol. 6, no. 46, eabc4516
DOI: 10.1126/sciadv.abc4516
  • Fig. 1 The cellobiohydrolase CbsA is associated with transitions to vascular pathogenic lifestyles in Gram-negative pathogens.

    (A) Highest-ranking associations between OG presence/absences and evolutionary transitions between vascular and nonvascular lifestyles in the Xanthomonadaceae. A genome-based SNP phylogeny is shown to the left, with strains from the same species condensed into clades. A heatmap summarizing, for each strain, the presence (black) or absence (white) of the two gene OGs, CbsA and hyp1, whose distributions are most strongly supported to be dependent on vascular lifestyle status (determined by model testing through the ranking of log Bayes factors; Materials and Methods) is shown to the right of the tree, followed by another heatmap indicating the classification of each strain as either vascular (blue), nonvascular (gray), or undetermined (beige) according to the literature (table S1). Additional figure details can be found in figs. S1 and S5. (B) Phylogenetic tree based on CbsA amino acid sequences from strains with whole-genome sequences found in (A), where branches on the tree are color-coded according to pathogenic lifestyle. To the right of each tip is a schematic depicting the neighborhood type in which that particular cbsA sequence is found, where the four possible neighborhood types are defined based on conserved synteny (indicated by color-coded gene models corresponding to specific OGs). Vascular bacteria have cbsA homologs located in type 1, 2, and 4 neighborhoods, while nonvascular bacteria have cbsA homologs found primarily in type 3 neighborhoods. Note that strains of the vascular pathogen X. campestris pv. campestris have two copies of cbsA located in either type 3 or 4 neighborhoods.

  • Fig. 2 Experimental gain and loss of CbsA facilitates transitions between vascular and nonvascular pathogenic lifestyles.

    (A) Addition of either cbsA from vascular X. translucens pv. translucens (Xtt) or cbsA from vascular X. oryzae pv. oryzae (Xoo) to nonvascular X. translucens pv. undulosa (Xtu) permits development of chlorotic lesions indicative of vascular disease on barley 21 days post-inoculation (dpi). (B) Corresponding vascular lesion lengths, with significant differences among treatments indicated by a to d (n = 6, P < 0.02). (C) Representative confocal images of vascular bundles downstream of leaf lesions on barley 12 dpi with GFP transformed strains demonstrate gain of vascular colonization by Xtu cbsAXoo. Green indicates bacterial cells expressing GFP; magenta indicates chlorophyll autofluorescence outlining nonvascular mesophyll cells; cyan indicates autofluorescence outlining xylem cell walls or phenylpropanoid accumulation in mesophyll cells. (D and E) Lesion lengths or incidence of nonvascular water-soaked lesions were quantified after barley leaf clipping 14 dpi with Xtt ∆cbsA. Bars in (E) represent percent leaves showing symptoms with dots included to display individual leaf lesion incidence. (F) Images of symptomatic barley leaves infected with Xtt and Xtt ∆cbsA, where water-soaked lesions are indicated, with black arrows indicating nonvascular symptom development.

  • Fig. 3 Repeated horizontal transfer, transposition, and gene loss events drive the distribution of cbsA in Gram-negative bacteria.

    (A) A 50% majority-rule consensus tree summarizing 81 conserved single-copy ortholog trees is shown to the left, with the names of the 75 individual isolates consolidated into relevant taxonomic groupings. Inferred HGT, transposition, and loss events are drawn and numbered on the tree and further described in (B). The matrix to the right of tree indicates the presence/absence of one of four distinct genomic neighborhood types (shaded/unshaded cells) in which cbsA homologs are found within a given genome (presence of cbsA indicated by an overlaid red arrow). Note that in many cases, all of the constituent genes making up a specific neighborhood are present in a given genome save for cbsA (indicated by the absence of an overlaid red arrow). cbsA homologs from X. albilineans and X. ampelinus were not found associated with a specific type of neighborhood, but were assigned to the type 2 neighborhood column based on the observation that their closest phylogenetic relatives are sequences in type 2 neighborhoods (see Fig. 1B). This tree has been lightly edited for viewing purposes by removing several taxa from outside the Xanthomonadales and can be viewed in its entirety in fig. S3. (B) Sequence of inferred evolutionary events numbered corresponding to (A). Genomic neighborhood types are represented by schematics, where gene models are color-coded according to OG. The color-coding of neighborhood types is consistent across both panels.

  • Fig. 4 The evolution of vascular and nonvascular pathogenesis in plant-associated Xanthomonas bacteria is driven by the gain and loss of cbsA.

    Our combined phenotypic and phylogenetic analyses support a model where vascular and nonvascular pathogenesis exist as two points on the same evolutionary continuum that is traversed by either the acquisition or loss of a single cellobiohydrolase, cbsA.

Supplementary Materials

  • Supplementary Materials

    Repeated gain and loss of a single gene modulates the evolution of vascular plant pathogen lifestyles

    Emile Gluck-Thaler, Aude Cerutti, Alvaro L. Perez-Quintero, Jules Butchacas, Ver�nica Roman-Reyna, Vishnu Narayanan Madhavan, Deepak Shantharaj, Marcus V. Merfa, C�line Pesce, Alain Jauneau, Taca Vancheva, Jillian M. Lang, Caitilyn Allen, Valerie Verdier, Lionel Gagnevin, Boris Szurek, Gregg T. Beckham, Leonardo De La Fuente, Hitendra Kumar Patel, Ramesh V. Sonti, Claude Bragard, Jan E. Leach, Laurent D. No�l, Jason C. Slot, Ralf Koebnik, Jonathan M. Jacobs

    Download Supplement

    The PDF file includes:

    • Supplementary Materials and Methods
    • Figs. S1 to S6 and S9
    • Legends for figs. S7 and S8
    • References

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article