Exploration of bacteriophages, endolysins, and cell wall binding domains of endolysins for control and rapid detection of bacteria

Abstract

B. cereus is an opportunistic human pathogen responsible for food poisoning and other nongastrointestinal infections. Due to the emergence of multidrug-resistant B. cereus strains, the demand for alternative therapeutic options is increasing. Bacteriophage-based therapies have been re-introduced as an additional tool, particularly in the war against multi-drug resistant pathogens. The enormous reservoir of novel genes within phages also provides potential resource for medical, molecular, and biotechnological applications. To understand genetic diversity of B. cereus phages and develop phage-based medical, molecular, and biotechnological tools, six B. cereus bacteriophages were isolated from various environmental samples, and characterized their genomes. Transmission electron microscopy analysis revealed that four phages (PBC1, PBC2, PBC4, and PBC5) were classified into the Siphoviridae and two phages (PBC6, PBC9) into the Myoviridae family. These phages generally have narrow host ranges within the B. cereus group species. Phages in the Siphoviridae showed relatively large ranges in genome size

PBC1 has the smallest genome of 41.2 kb, followed by 56.3 kb of PBC5, 80.6 kb of PBC4, and 168.7 kb of PBC2. The genomes of the highly related Myoviridae phages, PBC6 and PBC9, are 157.1 and 157.2 kb, respectively. The siphovirus PBC2 is closely related to the B. anthracis phage Tsamsa, while PBC4 showed high similarity to the B. cereus phage Basilisk over the entire genome. PBC1 and PBC5 seemed to be novel B. cereus phage as they showed a very low degree of nucleotide identity to the previously reported phages. Genome analysis revealed that the phage PBC1 is the only virulent phage that lacks virulence and antibiotic resistance genes among the isolated phages. Growth inhibition assay of B. cereus with liquid culture and boiled rice confirmed the strong lytic activity of PBC1. These results suggest that the virulent phage PBC1 could be a useful component of a phage cocktail to control B. cereus even with its exceptionally narrow host range as it can kill a strain of B. cereus that is not killed by other phages. Various phage-gene products showing antibacterial activity are reported be useful biocontrol agents. Bacteriophage endolysins are expressed at the end of the phage reproductive cycle, hydrolyzing the cell wall peptidoglycan to release virion progeny. Since endolysins are bactericidal enzymes with highly evolved specificity toward target bacteria, their potentials as antibacterial agents have been intensively studied. Moreover, modular nature of endolysin containing enzymatic active domains (EADs) and cell wall binding domains (CBDs) suggests that varieties of domains with unique activities can be derived from endolysins. Several endolysins from the isolated B. cereus phages were characterized and their functional domains, including EADs, CBDs, and a spore binding domain (SBD) were identified. Compared to the high host specificity of the B. cereus phages, their endolysins generally showed a much broader lytic spectrum, albeit limited to the genus Bacillus. The EAD of PBC1 endolysin when expressed alone also showed Bacillus-specific lytic activity, which was lower against the B. cereus group, but higher against the B. subtilis group than the full-length protein. The CBDs from B. cereus phage endolysins showed high specificity and binding capacity to the B. cereus group strains, proposing that they can be used as novel biological probes for B. cereus detection. The SBD of PBC2 endolysin specifically binds only B. cereus spores but not vegetative cells of B. cereus. The spores with disrupted exosporium nap layer displayed much higher binding of SBD, suggesting that the exosporium nap layer may not be the binding target of the PBC2_SBD. Acquisition of a novel CBD from a phage endolysin is time-consuming and labor-intensive. To address this issue, a simple method to identify CBDs from a sequenced bacterial genome was presented by employing homology search for phage lysin genes. A CBD from a genome of Clostridium perfringens was identified and confirmed that it is specific to C. perfringens cells. Because this method does not require phage isolation and phage genome sequencing, I think it could be a general approach for CBD identification from sequenced bacterial genomes. The target-specific binding activity of the CBDs of endolysins inspired me to exploit the feasibility of engineering CBD to detect multiple pathogens. To develop more efficient CBD-based detection tools, several issues should be considered. First, A CBD cocktail was provided for simultaneous detection of multiple pathogens by combining of CBDs recognizing B. cereus, C. perfringens, and Staphylococcus aureus, with different fluorescent markers. Second, because many CBDs show high affinity to its target cells in a strain-specific manner, they are not capable of detecting diverse environmental strains. To extend the binding range, dual CBD hybrids were created by combining two CBDs with complementary specificities. The resulting 1H4_CBD, which combines PBC1_CBD and PBC4_CBD via a helical linker, were capable of binding most B. cereus-group strains tested, and showed superior binding activity to its parental CBDs. This result demonstrated that the CBD can be tailored to recognize additional targets by adding other binding modules with suitable linkers. Lastly, with CBD-coated magnetic nanoclusters (CBD-MNCs), the significant portions of the viable bacterial cells could be separated from diluted suspensions within 30 min. More importantly, the CBD-MNCs showed better cell capture performance than an antibody-based approach, representing a potential of this method in developing CBD-based microbial diagnostics.