PMC:4996382 / 7234-14218 JSONTXT

{"project":"2_test","denotations":[{"id":"27600214-21906634-69478337","span":{"begin":953,"end":955},"obj":"21906634"},{"id":"27600214-21906634-69478338","span":{"begin":2587,"end":2589},"obj":"21906634"},{"id":"27600214-12527788-69478339","span":{"begin":2590,"end":2592},"obj":"12527788"},{"id":"27600214-22829736-69478340","span":{"begin":2593,"end":2595},"obj":"22829736"}],"text":"2. Experimental Section\n\n2.1. Bacillus Anthracis Genomes\nA set of eight Bacillus anthracis genomes was selected encompassing three levels: gapless chromosomes, scaffolds or contiguous. Genomes were downloaded from the NCBI Microbial genomes database (Table 1).\n\n2.2. Universal Fingerprinting Chip (UFC-13)\nAn oligonucleotide probe is a short piece of single-stranded DNA that is complementary to the target to be measured on the microarray. A set of 15,264 13-mer oligonucleotide sequences constitutes the UFC-13 chip that was used for hybridizing with all the Bacillus anthracis genomes strains selected. The UFC-13 includes: (a) a 35%–65% G + C content; (b) between-sequences differences have been maximized such that all sequences differ in at least three bases from each other; and (c) the sequences Tm ranges between 52 °C and 68 °C. Bacillus anthracis has an average of 5.5 Mb, and the size of the 13-mer probes is suitable for bacterial genomes [13].\nmicroarrays-04-00084-t001_Table 1 Table 1 Bacillus anthracis genomes used in this study. * Related GenBank Project (Accession).\n\n2.3. Virtual Hybridization\nVH software was used to calculate in silico genome fingerprints based on the complementarity between the probes and the genome. VH can also calculate and simulate various thermodynamic parameters such as Gibbs free energy (ΔG°), number of mismatches and melting temperature (Tm).\nWe have used the set of 13-mer sized probes in previous studies and found this probe size to be suitable for studying bacterial genomes (0.5–10 Mbp). The virtual hybridization reaction between the UFC-13 probes and the Bacillus anthracis genomes is first performed to identify possible sites of DNA duplex formation. Then, a computer simulation of the hybridization reaction between probes in the array and the target sequences (genomes) is conducted to predict the hybridization patterns that can be possibly obtained under the experimental conditions set. VH is carried out in two steps. The first step aims to finding the sites where DNA duplex formation (probe-target) might most likely occur. Such sites are identified by comparing the probes sequence with that of the target, with basis on their bases complementarity; these sites are labeled “potential hybridization sites”. In the second step, the free energy between the probe and the potential hybridization sites is calculated. VH accurately maps the position of each probe in the genome, identifying specific probes for each Bacillus anthracis strain, and yielding a signal or the formation of a spot when the duplex is formed [13,15,16].\n\n2.4. Virtual Hybridization by Direct and Extended Methods\nVH utilizes two methods: (a) the direct method identifies all the sites where virtual hybridization with the genome might potentially occur and then, using the cut-off value, identifies the sites with high probability for heteroduplex formation (Figure 1); (b) the extended method also identifies the sites where virtual hybridization between the genome and the UFC-13 might potentially occur, but signals corresponding to non-conserved sequences are discarded to leave only those corresponding to virtual hybridization with homologous sites. This method increases the alignment between the probe and the site in the genome, by adding four nucleotides onto the right and left ends of the probe (4 + 13 + 4), with the sequences present in the target DNA, allowing only one difference between them, thus ensuring that their coincidence is not random, and can, therefore, be considered as conserved (Figure 2).\nFigure 1 Virtual hybridization (VH)—Direct method. Illustrative example of four probes being used to predict the sites where virtual hybridization with the target genome (represented by the blue line) might potentially occur. Then, VH software estimates the ΔG° value for each of the probes and identifies the high-probability sites.\n\n2.5. Genomic Fingerprints\nVH data were used to create a unique genomic fingerprint specific for each bacterial strain. UFA (Universal Fingerprint Analysis) generated an in silico microarray for each bacterium. The software output shows the number of columns and rows (spots) on the microarray.\nFigure 2 Virtual hybridization (VH)—Extended method. Illustrative example of four probes used to predict sites where virtual hybridization with the target genome (represented by the blue line) might potentially occur. The VH software estimates the ΔG° value for each of the probes, adds four bases at the ends of the 13-nucleotide probes to yield 21-bases long (4 + 13 + 4) segments, and identifies whether the sequences are identical. It is statistically highly unlikely for two sequences of this length to have a high degree of similarity by chance. The genomic fingerprint obtained shows the sites where virtual hybridization with the genome took place and identifies the positions where a DNA heteroduplex was formed. The fingerprint obtained can then be compared to other fingerprints to identify spots that are specific for individual bacteria. Visual microarrays render an image of the genomic fingerprint of each Bacillus anthracis. This image represents an in silico DNA microarray for a given organism, along with the specific probes used in hybridization experiments. This tool shows the set of 15,264 probes on a microarray as spots, color‑coded to identify those probes that hybridized with a particular target.\nThe microarray_pic software provides a very useful tool to display virtual hybridization patterns (fingerprint) graphically. This graphical representation shows the probes-to-target signals where a duplex was formed. Sites with high probability of virtual hybridization are shown with a green- or red-colored spot. Some probes can hybridize at multiple sites in the genome. In addition, two different tracks can be overlapped. The overlap shows, in yellow color, those probes that are shared by the two organisms; probes that are specific to one of the organisms are shown in green while those specific to the other organism are shown in red.\nIn silico genomic fingerprints were obtained for eight Bacillus anthracis strains. The fingerprints were obtained with the UFC-13 by virtual hybridization with 15,264 probes. Thermodynamic parameters used were: 1 mismatch and a ΔG° cutoff range of −19.53 to −11.67 kcal/mol. The method also identified those probes that are highly specific and highly potential for each Bacillus strain. The output file provides specific information for each microarray: (a) probe number; (b) probe ID; (c) position in the target sequence; (d) target sequence; and (e) ΔG°.\n\n2.6. Bacillus Anthracis Fingerprint Tree\nFingerprints obtained for the Bacillus anthracis strains were compared with each other and pairwise distances for all the possible fingerprint pairs were calculated. A taxonomic tree was then constructed from the pairwise fingerprint distances matrix, using the Neighbor-Joining algorithm in the PHYLIP 3.61 software.\n\n"}