Removal of group B streptococci colonizing the vagina and oropharynx of mice with a bacteriophage lytic enzyme

Abstract

Group B streptococci (GBS) are the leading cause of neonatal meningitis and sepsis worldwide. The current treatment strategy is limited to intrapartum antibiotic prophylaxis in pregnant women to prevent early-onset neonatal diseases, but considering the potential for antibiotic resistance, the risk of losing control over the disease is high. To approach this problem, we have developed a bacteriophage (phage) lytic enzyme to remove colonizing GBS. Bacteriophage muralytic enzymes, termed lysins, are highly evolved molecules designed to degrade the cell wall of host bacteria to release phage particles from the bacterial cytoplasm. Several different lysins have been developed to specifically kill bacterial pathogens both on mucosal surfaces and in blood and represent a novel approach to control infection. A lysin cloned from a phage infecting GBS was found to contain two putative catalytic domains and one putative binding domain, which is similar to the domain organization of some staphylococcal phage lysins. The lysin (named PlyGBS) was recombinantly expressed in Escherichia coli, and purified PlyGBS efficiently killed all tested GBS serotypes in vitro. In a mouse model, a single dose of PlyGBS significantly reduced bacterial colonization in both the vagina and oropharynx. As an alternative strategy for intrapartum antibiotic prophylaxis, this approach may be used to reduce vaginal GBS colonization in pregnant women before delivery or to decontaminate newborns, thus reducing the incidence of GBS-associated neonatal meningitis and sepsis.

FIG. 1.

Schematic diagram of domain organization in PlyGBS. Based on sequence analysis, PlyGBS contains two catalytic domains: the N-terminal region has 27% identity to the first 94 amino acids of the endopeptidase domain of staphylococcal phage 187 lysin (Ply187), and the central region of the PlyGBS displays 46% identity to the muramidase domain of pneumococcal phage Cp-1 lysin (Cpl-1). The C terminus of PlyGBS is a putative cell wall binding domain. The domain organization in PlyGBS resembles that of some staphylococcal phage lysins, which have two or more catalytic domains, such as Ply187 containing endopeptidase, amidase, and glucosaminidase domains.

FIG. 2.

Coomassie blue-stained SDS-PAGE of purified PlyGBS. The molecular mass of the protein ladder is presented in kilodaltons.

FIG. 3.

PlyGBS activity against GBS in vitro. GBS NCTC 11237 cells were resuspended in 40 mM phosphate buffer (pH 5.0) and mixed in 96-well microtiter plates with 40 U of PlyGBS. The OD₆₀₀ (solid triangles) was monitored by a spectrophotometer. To check the cell viability (solid squares), the cells were centrifuged, diluted, and plated on THY agar plates for enumeration. Control experiments (open circles) with the addition of phosphate buffer (pH 5.0) were performed under the same conditions.

FIG. 4.

Lysin activity against different bacteria in vitro. The cells were exposed to 15 U of PlyGBS for 60 min, and the decrease of OD₆₀₀ was monitored by a spectrophotometer. The activity of lysin was expressed as the initial velocity of the decrease in absorbance over time (−mOD₆₀₀/min). Error bars show the standard error of the mean.

FIG. 5.

Visualization of the muralytic effect of lysin on GBS cells by phase-contrast and electron microscopy. (a and b) Phase-contrast microscopy of cells treated with buffer (a) or PlyGBS (b) for 15 min. (c to e) Thin-section transmission electron micrographs (magnification, ×65,000) of cells treated with PlyGBS show the extrusion (c) and rupture (d) of the cytoplasmic membrane which occurred at the septum region, resulting in “ghost” cells (e) after the loss of cytoplasmic contents.

FIG. 6.

Control of GBS colonization in mice with PlyGBS. (a) Mice were colonized with GBS vaginally after synchronization with β-estradiol valerate. One day after GBS inoculation, two groups of mice were treated vaginally with either buffer (n = 15) or PlyGBS (n = 15). Each mouse was swabbed vaginally before treatment (0-h samples) and after treatment at 2- to 4-h intervals (2- and 4-h samples). (b) Mice were colonized with GBS in their upper respiratory tract, and 1 day later they were treated with either buffer (n = 18) or PlyGBS (n = 20). Each mouse was swabbed before (0 h) and after treatment (2 and 24 h). After transformed into log₁₀, the colony counts of oropharynx samples were averaged for each time interval in the same group and plotted. Error bars represent the standard error of the mean.