Effects of DNA size on transformation and recombination efficiencies in Xylella fastidiosa

Abstract

Horizontally transferred DNA acquired through transformation and recombination has the potential to contribute to the diversity and evolution of naturally competent bacteria. However, many different factors affect the efficiency with which DNA can be transformed and recombined. In this study, we determined how the size of both homologous and nonhomologous regions affects transformation and recombination efficiencies in Xylella fastidiosa, a naturally competent generalist pathogen responsible for many emerging plant diseases. Our experimental data indicate that 96 bp of flanking homology is sufficient to initiate recombination, with recombination efficiencies increasing exponentially with the size of the homologous flanking region up to 1 kb. Recombination efficiencies also decreased with the size of the nonhomologous insert, with no recombination detected when 6 kb of nonhomologous DNA was flanked on either side by 1 kb of homologous sequences. Upon analyzing sequenced X. fastidiosa subsp. fastidiosa genomes for evidence of allele conversion, we estimated the mean size of recombination events to be 1,906 bp, with each event modifying, on average, 1.79% of the nucleotides in the recombined region. There is increasing evidence that horizontally acquired genes significantly affect the genetic diversity of X. fastidiosa, and DNA acquired through natural transformation could be a prominent mode of this horizontal transfer.

Fig 1

Recombination efficiencies of plasmids with flanking regions varying from 96 to 4,000 bp (A) and from 26 to 1,000 bp (B; note that the x axis in panel B is on a log scale). Solid lines and squares show the recombination efficiencies (number of recombinants per total number of cells present) after normalization for the total number of transforming DNA units present. Dotted lines and empty diamonds show overall recombination efficiencies. Recombination efficiencies peaked with approximately 1,000 bp of flanking region and then plateaued. Recombinants were recovered with as few as 96 bp of homology, but rates were essentially 0. Images on the right depict the regions cloned into the vector backbone pGEM-5zf(+) to create plasmids with various flanking regions. Homologous DNA was amplified from the rpfF region of X. fastidiosa.

Fig 2

Recombination efficiencies of plasmids with nonhomologous inserts of different lengths ranging from 1 to 6 kb. Solid lines and squares show the recombination efficiencies (number of recombinants per total number of cells present) after normalization for the total number of transforming DNA units present. Dotted lines and empty diamonds show the overall recombination efficiencies. No recombinants were recovered when the total insert length was 6 kb. The image on the right depicts the region cloned into the vector backbone pGEM-5zf(+) to create the plasmids. The flanking region size was kept constant at approximately 1 kb on either side. The nonhomologous insert region consisted of a kanamycin resistance cassette (approximately 1.1 kb) and 0 to 5 kb of DNA amplified from a plant virus with no regions of homology to the X. fastidiosa genome (the total insert size for pS2 is approximately 1 kb, that for pS2-1k is approximately 2 kb, etc.). Arrows over the homologous DNA regions show the locations of primers rpfF-fwd and rpfF-rev, which were used to screen antibiotic-resistant colonies for proper insertion of the kanamycin resistance cassette and nonhomologous DNA.