Molecular Identification and Phylogenetic Analysis of Multidrug-resistant Bacteria using 16S rDNA Sequencing

1department of clinical laboratories, Faculty of applied Medical Sciences, Taif university, Saudi arabia. 2department of nursing, Faculty of applied Medical Sciences, Taif university, Saudi arabia. 3genetics department, Faculty of agriculture, Kafrelsheikh univ., 33516, Kafr El-Sheikh, Egypt 4department of biology, Faculty of Science, Taif university, Saudi arabia. 5department of genetics, Faculty of agriculture, Minufiya university, Egypt. 6department of cell biology, national research center, dokki, giza, Egypt. 7department of genetics, Faculty of agriculture, cairo university, Egypt.

A developed multidrug-resistant (MDR) hospital habitat microorganism for specific uses is part of understanding the mechanisms of the transfer of antibiotic resistance genes among different bacterial species. Isolation and identification of MDR microorganisms is challenging because these microorganisms exhibit highly specific characteristics that make them grow in their natural environments, such that they cannot be cultured using standard laboratory methods (Vijayan et al., 2012). Competition among microbes for space and nutrients in the hospital environment represents a powerful selection pressure that can confer enhanced resistance of MDR microorganisms to several drugs (Jeganathan et al., 2013). Biogeographical relationships among MDR bacteria can be structurally analyzed via 16S rDNA sequencing, which has become the cornerstone in microbial ecology for investigating bacterial community structures in various environments (Hur and Chun, 2004;Amann, 2000). The 16S rDNA sequence has properties that make it highly suitable as a universal phylogenetic marker. Furthermore, the 16S rDNA gene sequencing is a useful method for bacterial classification, in which the nucleotide sequences of the this region are determined and compared with sequences available from databases to yield homology matches, thereby allowing bacterial identification of the target samples (Clarridge 2004;Uyeno et al., 2004;Salman et al., 2012;Sabir et al., 2013). In the present study, MDR bacterial strains were isolated from patient samples collected from different hospitals in Taif City and grown in different culture media. Samples were classified into three phylogenetic groups based on 16S rDNA analysis (da Silva et al. 2013). Isolation and identification of potential strains of Streptomyces sp., Bacillus sp., Flavobacterium sp., and Pseudomonas sp. from patients has been previously conducted (Anand et al., 2011). Vibrio, Staphylococcus, and Bacillus are pathogenic MDR bacteria that cause infection in humans and harbors strains with antagonistic activity (Adiga et al., 2010;Gram et al., 2010;. In the present study, a total of 30 MDR bacterial strains were identified using 16S rDNA sequence analysis and bioinformatics analysis. Evolutionary and phylogenetic relationships among isolated strains were also studied. The effectiveness of this method was demonstrated by comparison with various bacterial communities collected from the GenBank database.

sample collection
A total of 200 clinical samples of urine and stool swabs were collected from hospitalized patients in Taif hospitals, Saudi Arabia within a sixmonth period. Hospitalized patients were randomly selected and asked to undergo screening for nasal carriage of antibiotic resistance strains. Clinical samples were obtained under aseptic conditions and transported immediately to the microbial genetics laboratory at the Biotechnology and Genetic Engineering Unit at Taif University. isolation and purification of clinical bacterial isolates Sterile dry swabs were used to streak bacterial samples onto sterile Petri dishes containing nutrient agar media (BioLife, USA). Inoculated dishes were incubated at 28 °C for 48 h. Single colonies were obtained using sterile inoculation needles, and slant cultures of nutrient agar media were prepared.

Genomic dna extraction and PCr amplification
Genomic DNA was extracted from the cell pellets of all MDR isolates using a DNA extraction kit (Jena Bioscience, Germany) according to the manufacturer's instructions. For each isolate, a single DNA fragment (about 1200 bp) of the 16S rDNA gene was amplified according to previously described methods (Hassan and Ismail, 2014). Fragments were purified using QIAquick PCR purification kit (QIAGEN, Valencia, CA, USA) and sequenced on an Applied Bioscience 3130 DNA analyzer (Applied Biosystems, USA). Sequence reads were edited and assembled using the DNASTAR software (Lasergene, Madison, WI, USA). BLAST searches were conducted using the NCBI server (http://www.ncbi.nlm.nih.gov/blast/ Blast.cgi).

Bioinformatic analysis
The 30 sequences were analyzed using the BLAST query tool (http://blast.ncbi.nlm. nih.gov) to determine non-chance sequence similarities. BLAST queries were restricted to the 16S ribosomal RNA database. Models (XM/XP) and uncultured/environmental samples were also filtered out to obtain more reliable results. Each sequence was analyzed using blast, and the blast hit with the lowest expect-value (e-value, which indicates the number of non-chance alignments) was selected ( Table 1). The stringent e-value value cutoff of 1e -6 was selected to ensure that significant blast outputs were obtained.

Blast analysis and sequence variation
Results of NCBI BLAST analysis of the 30 sequences are shown in Table (1). Query sequences were evaluated and compared with the database sequences based on the following criteria: query coverage, e-value, and percent identity. Query coverage represents the percentage of query sequence length that is included in the alignment. The e-value is a parameter that represents the number of chance alignment and directly depends on the size of the search space (database), such that a smaller the database has higher likelihood of containing the sequence of interest. Search queries were restricted to the 16S ribosomal RNA database. Models (XM/XP) and uncultured/environmental samples were filtered out. The percent identity refers to the percentage of residues in the aligned sequences that match at the same position. BLAST analysis classified the 30 isolates into the following 13 different genera: Bacillus, Caldimonas, Citrobacter, Enterococcus, Escherichia, Klebsiella, Lactobacillus, Lactococcus, Leuconostoc, Micrococcus, Pantoea, Pseudomonas, and Staphylococcus.
Zero e-values were obtained for all 30 queries, which indicated that all alignments were significant. The percent query coverage values ranged from 76% to 100%. In addition, we observed high percent identities, which ranged from 84% to 99%. Table 2 shows the description, sequence length, and GC content for each isolate. Sequence length was highly variable among the 30 isolates and within genera. The shortest sequence was 728 bp, which corresponds to S. aureus, whereas B. cereus had the longest sequence of 1,219 bp. On the other hand, GC content ranged from 50% to 57%. The base frequencies of all 30 isolates are presented in Figure 1. Only the two M. yunnanensis isolates showed significant variation by having the lowest adenine and cytosine contents.

Polymorphism and genetic diversity among species
The 30 sequences were aligned using both Clustalx (Larkin et al., 2007) and MAFFT (Katoh et al., 2002) build in the ape package (Paradis et al., 2004). The alignment results of the two software were found to be identical, demonstrating the robustness of the alignment methods. The aligned sequences had the same length of 1,219 bp. Sequence analysis revealed a total of 285 polymorphic sites (excluding sites with gaps or missing data), which comprised 116 invariable (monomorphic) sites and 169 variable (polymorphic) sites. Of these, 36 sites were singletons (non-informative) and 133 were parsimony-informative sites. Results of DNA polymorphism analysis of the 30 isolates sequences are shown in Table 3. The nucleotide diversity (ð) was 0.17, while the number of haplotypes (h) was 23. The haplotype diversity (H d ) was 0.94, with a standard deviation of 0.04. Only four genomic conserved regions were detected along the 1480-bp region ( Table 4). The shortest sequence was 106 bp, while the longest was 307 bp. These conserved regions ranged from 606 and 1453 bp in length, and no conserved regions were detected in almost first 600 bp. Furthermore, the two sequences contained overlapping regions.

Phylogenetic analysis and evolutionary relationships
The evolutionary histories of 30 taxa were inferred using the neighbor-joining method (NJ) (Saitou & Nei, 1987) (Figure 2). The constructed phylogenetic tree clearly resolved the relationships among the 30 sequences, and most of bootstrap supports were high. The optimal tree with the sum of branch length = 1.44214152 is shown. The tree is drawn to scale, with branch lengths having the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were calculated using the Maximum Composite Likelihood method (Tamura et al., 2004) and were expressed as the number of base substitutions per site. All positions containing gaps and missing data were eliminated. The final dataset comprised a total of 285 positions. The 30 sequences were classified into three large clades

disCussion
Transfer of MDR genes among different bacterial species leads to increased biological and genetic diversity. Studying MDR bacteria is important for elucidating the mechanisms of antibiotic resistance among different bacterial species. Despite very high microbial diversity, the lack of culture protocols for the microbes that are most abundant in the environment significantly limits bio-discovery research. Maintaining bacterial cultures is a critical step for developing applications of various microbes. Results of the present study clearly indicated that bacterial culture combined with 16S rDNA sequence analysis provides a novel means of comparing multiple bacterial community structures. We performed bioinformatics analysis to compare multiple bacterial communities and determine the evolutionary and phylogenetic relationships among our isolates. The following strains were identified from the isolates: 3 B. cereus, 1 B. subtilis, 1 B.  tequilensis, 1 C. manganoxidans, 1 C.  Analysis revealed monomorphic and polymorphic sites in some of the target strains, and parsimony-informative sites were identified in B. cereus and B. thuringensis. Nucleotide diversity (ð) is defined as the average number of nucleotide differences per site between two sequences (Nei, 1987). The ð estimates in the present study not only indicated that all the analyzed sequences not equally diverse site-wise, but also showed high nucleotide diversity, which ranged from 0.17 to 0.94. Our findings were consistent with those reported in previous study (Goodfellow and Bul, 2003) because the average nucleotide diversity (ð) showed high variability in terms of nucleotide diversity. In conclusion, our study indicated that the observed variability in sequences of our target bacterial strains can be attributed to antibiotic resistance and gene transfer among bacterial strains, which emerged because of the hospital environment.

aCknowledGeMents
We gratefully thank all the patients who were enrolled in this study. We thank the technical staff, the nursing, medical assistant and medical staff at the participating hospitals for their help with recruitment of the patients for the study.