Chromaspro 1

The nucleotide sequences of each gene obtained in this study were edited using the software ChromasPro 1.5 (Technelysium, Pty, Ltd., Brisbane, Australia) and were aligned with each other and with reference sequences from GenBank using ClustalX 2.0.6. Alignment adjustments were made manually to remove artificial gaps using BioEdit. Phylogenetic analyses were performed using the software MEGA5 [48] (link). Neighbour joining (NJ) trees were constructed. All ambiguous positions were removed for each sequence pair. The reliability of branches in trees was assessed using the bootstrap analysis with 1000 pseudo-replicates, with values above 50% reported. Phylograms were drawn using the MEGA5 and were manually adjusted using CorelDrawX5 (Ottawa, Canada). ITS and SSU sequences have been deposited in GenBank under the accession numbers KJ469967–KJ469979, respectively.

DNA sequences were edited with ChromasPro1.5 2003–2009 (Technelysium Pty Ltd, Helensvale, Australia) and aligned using ClustalW (http://workbench.sdsc.edu; Bioinformatics and Computational Biology group, Dept. Bioengineering, UC San Diego, CA, USA). The sequences used in phylogenetic analysis were chosen from the highest match based on BlastN result in GenBank against the four RKN species recovered from this study. The model of base substitution in the DNA sequence data was evaluated using MODELTEST version 3.06 [29 (link)]. The Akaike-supported model [30 ], the proportion of invariable sites, and the gamma distribution shape parameters and substitution rates were used in phylogenetic analyses using DNA sequence data. Bayesian analysis was performed to confirm the tree topology for each gene separately using MrBayes 3.1.0 [31 ], running the chain for 1,000,000 generations and setting the ‘burnin’ at 1,000. Markov Chain Monte Carlo (MCMC) methods were used within a Bayesian framework to estimate the posterior probabilities (pp) of the phylogenetic trees [32 ] using the 50% majority-rule. The λ2 test for homogeneity of base frequencies and phylogenetic trees was performed using PAUP* version 4.0 (Sinauer Associates, Inc. Publishers, Sunderland, MA, USA).

Genomic DNA was extracted from whole blood using QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany). All six exons and exon-intron boundaries and 5′ and 3′ UTR regions of AGPAT2 gene were amplified by polymerase-chain reaction (primer sequences listed in Table 1). PCR products were directly sequenced (Macrogen, Korea). The resulting sequences were compared with the published reference AGPAT2 transcript ID (ENST 00000371696) using ChromasPro 1.5 (Technelysium, Australia). Sequence variations were named in agreement with the recommendations of the Human Genome Variation Society [11] (link). Intron mutation analysis was performed with Human Splice Finder engine (http://www.umd.be/HSF/) that predicts the likelihood of splicing abnormalities. Natural splicing sites (5′and 3′), branching points and exon splice enhancers are assigned a consensus values. A delta Consensus Value (ΔCV) of the splice sites strength >10% is likely to determine splicing variations [12] (link).

For each plastid marker, both forward and reverse strands were assembled using ChromasPro 1.7.5 (Technelysium Pty Ltd, Australia) and haplotype sequences were deposited at GenBank (available at https://www.ncbi.nlm.nih.gov/genbank/) under numbers: C. elegansJQ082471.1, JQ082474.1, KM982182.1, KM982262.1; P. mantiqueirensisAY772921.1, DQ792340.1. DNA sequences were aligned using MEGA7 software (Kumar et al. 2016 (link)) with the ClustalW algorithm, and manually edited when necessary. Haplotype (h) and nucleotide (π) diversity indexes and the number of variable sites were calculated with Arlequin 3.5.1.2 (Excoffier and Lischer 2010 (link)). The two plastid intergenic spacers were concatenated and treated as a single sequence for all analyses. The numbers of variable and informative sites in the manually edited alignment were obtained from Mega7 software. Haplotypes were identified via DNAsp 5.10.01 (Rozas et al. 2003 (link)).

The nucleotide sequences of each gene obtained in this study were edited using the ChromasPro 1.7.5 software (Technelysium, Pty, Ltd.), manually edited, and aligned with each other and with reference sequences from GenBank using MAFFT version 7 online server with automatic selection of alignment mode (http://mafft.cbrc.jp/alignment/software/). Phylogenetic analyses were performed and best DNA/Protein phylogeny models were selected using the MEGA6 software [73 (link), 74 (link)]. Phylogenetic trees were inferred by the i) neighbor-joining (NJ), ii) maximum likelihood (ML), and iii) maximum parsimony (MP) method. Bootstrap support for branching was based on 1000 replications. Neighbor-joining phylograms were edited for style using CorelDrawX7. Sequences have been deposited in GenBank under the accession numbers KR090615-KR090632 and KT731193-KT731212.

MLST was carried out as described previously using 44 strains representing the three most commonly isolated species: A.baumannii (Diancourt et al. 2010 (link)), P.aeruginosa (Curran et al. 2004 (link)), and S.maltophilia (Kaiser et al. 2009 (link)). Sequencing was performed by Genomed SA (Warsaw, Poland), and the resulting sequences were analyzed using ChromasPro 1.4 software (Technelysium Pty Ltd., Australia). MLST data were then compared with the database (http://pubmlst.org) to identify sequence types.

Multi-locus sequence typing (MLST) was performed as described by Diancourt et al. [32 (link)] with a Pasteur scheme. Sequencing of housekeeping genes cpn60, fusA, gltA, pyrG, recA, rplB, rpoB was performed by Genomed SA (Warsaw, Poland), and the resulting sequences were analyzed using the ChromasPro 1.4 software (Technelysium Pty Ltd., South Brisbane, QLD, Australia). Sequence types (STs) were determined by comparison to the PubMLST database [33 (link)].