Assessment of Multilocus Sequence Analysis (MLSA) for Identification of Candidatus Liberibacter Solanacearum from Different Host Plants in Spain

Abstract

Liberibacter is a bacterial group causing different diseases and disorders in plants. Among liberibacters, Candidatus Liberibacter solanaceraum (CLso) produces disorders in several species mainly within Apiaceae and Solanaceae families. CLso isolates are usually grouped in defined haplotypes according to single nucleotide polymorphisms in genes associated with ribosomal elements. In order to characterize more precisely isolates of CLso identified in potato in Spain, a Multilocus Sequence Analysis (MLSA) was applied. This methodology was validated by a complete analysis of ten housekeeping genes that showed an absence of positive selection and a nearly neutral mechanism for their evolution. Most of the analysis performed with single housekeeping genes, as well as MLSA, grouped together isolates of CLso detected in potato crops in Spain within the haplotype E, undistinguishable from those infecting carrots, parsnips or celery. Moreover, the information from these housekeeping genes was used to estimate the evolutionary divergence among the different CLso by using the concatenated sequences of the genes assayed. Data obtained on the divergence among CLso haplotypes support the hypothesis of evolutionary events connected with different hosts, in different geographic areas, and possibly associated with different vectors. Our results demonstrate the absence in Spain of CLso isolates molecularly classified as haplotypes A and B, traditionally considered causal agents of zebra chip in potato, as well as the uncertain possibility of the present haplotype to produce major disease outbreaks in potato that may depend on many factors that should be further evaluated in future works.

Keywords: HLB; Liberibacter; MLSA; carrot; celery; citrus; parsnip; potato; zebra chip.

Figure 1

Phylogenetic tree of Liberibacter strains based on partial sequences of the 50S ribosomal proteins L10/L12 genes (rplJ/rplL). The tree was based on nucleotide sequences from 39 isolates and 453 nt positions. Analysis was performed using the Maximum Likelihood method and the Tamura-Nei model [28]. Bootstrap values (1000 replications) are shown at the branch points. Grey circles indicate those samples sequenced in this work. Phylogenetic analyses were conducted in MEGA X [27].

Figure 2

Example of phylogenetic tree of representative Liberibacter strains based on partial sequence of adenosine kinase gene (adk). The tree was based on nucleotide sequences from 30 isolates and 208 nt positions. Analysis was performed using the Maximum Likelihood method and the Tamura-Nei model [28]. Bootstrap values (1000 replications) are shown at the branch points. Grey circles indicate those sequenced in this work. Phylogenetic analyses were conducted in MEGA X [27].

Figure 3

Taxonomic resolution based on the number of base differences per site (p-distance) between sequences from the Liberibacter (a,c) and CLso (b,d) phylogroups using individual genes (a,b) or multilocus sequence analysis (MLSA) approaches (c,d) by different concatenated gene combinations: adk-mutS-icdA (A), adk-dnaG-icdA (B), adk-dnaG-recA (C), adk-mutS-recA (D,), adk-icdA (E), adk-recA (F), mutS-icdA (G), mutS-recA (H). The evolutionary history was inferred using the Maximum Likelihood method and the Tamura-Nei model [28]. Bootstrap values (1000 replications). Analysis was conducted in MEGA X [27].

Figure 4

Phylogenetic tree of representative isolates of CLso, CLas, CLaf, and Lcr based on partial sequence of adk (a), dnaG (b), fumC (c), grpE (d), icdA (e), metG (f), mutS (g), purA (h), recA (i) and gyrB (j). Analysis was performed by using the Maximum Likelihood method and the Tamura-Nei model [28]. Bootstrap values (1000 replications) are shown at the branch points.

Figure 5

Phylogenetic tree of representative isolates of CLso, CLas, CLam, CLaf and Lcr based on concatenated sequences obtained from adk, dnaG, fumC, grpE, icdA, metG, mutS, purA, recA and gyrB. The evolutionary history was inferred using the Maximum Likelihood method and the Tamura-Nei model [28]. Bootstrap values (1000 replications) are shown at the branch points. A total of 3967 nt positions in the final dataset were analysed in MEGA X [27].

Figure 6

Phylogenetic trees of CLso, CLas, CLaf and Lcr based on concatenated sequences obtained from adk-mutS-icdA (a), adk-dnaG-icdA (b), adk-dnaG-recA (c), adk-mutS-recA (d), adk-icdA (e), adk-recA (f), mutS-icdA (g), and mutS-recA (h). The evolutionary history was inferred using the Maximum Likelihood method and the Tamura-Nei model [28]. Bootstrap values (1000 replications) are shown at the branch points. A total of 1037 (adk-mutS-icdA), 1012 (adk-dnaG-icdA), 1083 (adk-dnaG-recA), 1108 (adk-mutS-recA), 602 (adk-icdA), 673 (adk-recA), 830 (mutS-icdA), and 901 (mutS-recA) positions in the final dataset were analysed in MEGA X [27].

Figure 7

Timetree inferred by applying the RelTime method [31,32] to the user-supplied phylogenetic tree whose branch lengths were calculated using the Maximum Likelihood (ML) method and the Tamura-Nei substitution model [28]. The time tree was computed using one calibration constraint (147 Myr between CLas/CLaf). This analysis involved nine nucleotide sequences and BT1 of L. crescens was used as outgroup. There were a total of 3933 positions in the final dataset. Evolutionary analyses were conducted in MEGA X [27].