Antibiotic resistance in Red Sea brine pools

Antibiotic resistance (AR) is a complex problem with a global clinical impact. However, did this phenomenon begin in conjunction with the medical use of antibiotics or is it older? Studying AR in pristine environments could answer this question. Being devoid of anthropogenic impact, makes Red Sea brine pools ideal targets for the study. Besides, the extremophilic nature of these pools, particularly Atlantis II Deep (ATIID) allows mining for novel thermostable AR genes, which could provide better understanding of AR evolution and enrich the thermophilic selection marker gene repertoire. Here, we initially validated commonly used AR detection methods, then analyzed antibiotic resistance in four brine pools (Atlantis II, Discovery, Kebrit and Chain Deeps) in addition to a brine-influenced site. Publicly available metagenomes with varying degrees of human impact were also included. Analysis was carried out by alignment of Roche-454 metagenomic reads using BLASTX to antibiotic resistant polypeptides contained in the Comprehensive Antibiotic Resistance Database (CARD, http://arpcard.mcmaster.ca/). Reads were assigned to the best hit with more than 90% identity over at least 25 amino acids. Reads aligning to genes, whose resistance is conferred by mutation, were screened to pinpoint these mutations. The analysis also involved determination of the abundance and diversity of three different mobile genetic elements (MGEs), namely plasmids, insertion sequences and integrons. Moreover, two open reading frames (ORFs), identified from ATIID through BLASTX alignment to CARD, were synthesized, cloned and expressed. Results showed a caveat in the current AR detection methods represented in the annotation of mutation-generated resistance genes. AR analysis in brine pools and publicly available metagenomes detected antibiotic resistance genes in 21 out 32 samples (65.6 %). Several genes were identified, conferring resistance to different classes of antibiotics, including beta-lactams, rifampin, fluoroquinolones, macrolides and aminoglycosides. Analysis of MGEs showed statistically significant correlation between AR abundance and the abundance of both plasmids and integrons. Interestingly, the abundance of MGEs, particularly insertion sequences showed strong association with extreme conditions in ATIID. On the other hand, the expression of synthesized ORFs, which putatively coded for a class A beta-lactamase (ABL) and a 3\u27-aminoglycoside phosphotransferase (APH(3\u27)), confirmed the annotation of both through enzyme assays, while only (APH(3\u27)) showed resistance in Escherichia coli. Remarkably, (APH(3\u27)) proved to be thermostable (Tm = 61.7 °C and ~40% residual activity after 30 min at 65 °C). In contrast, ABL was not as thermostable; Tm = 43 °C. In conclusion, we rectified the current AR detection methods through accurate account for resistance-causing mutations. We also provide a new evidence that environmental microorganisms represent a reservoir for AR genes. In addition, we shed light on the role of MGEs in the spread of antibiotic resistance and highlight the potential role of insertion sequences in the evolution of extremophiles. We also discovered two novel antibiotic resistance enzymes with potential application as thermophilic selection markers