Use of genetically engineered phage to deliver antimicrobial agents to bacteria: an alternative therapy for treatment of bacterial infections

Abstract

The emergence and increasing prevalence of multidrug-resistant bacterial pathogens emphasizes the need for new and innovative antimicrobial strategies. Lytic phages, which kill their host following amplification and release of progeny phage into the environment, may offer an alternative strategy for combating bacterial infections. In this study, however, we describe the use of a nonlytic phage to specifically target and deliver DNA encoding bactericidal proteins to bacteria. To test the concept of using phage as a lethal-agent delivery vehicle, we used the M13 phagemid system and the addiction toxins Gef and ChpBK. Phage delivery of lethal-agent phagemids reduced target bacterial numbers by several orders of magnitude in vitro and in a bacteremic mouse model of infection. Given the powerful genetic engineering tools available and the present knowledge in phage biology, this technology may have potential use in antimicrobial therapies and DNA vaccine development.

FIG. 1.

Kinetics of gef and chpBK induction. Cells of E. coli ERAPlacI harboring (A) pUPRIP, (B) pGef, and (C) pChpBK were grown overnight at 37°C and then diluted 1:100 into fresh LB medium containing appropriate antibiotics. At early log phase (OD₆₀₀ of 0.3 to 0.4), the culture was divided equally (indicated by arrow) and incubated in the presence (•) or absence (○) of 1 mM IPTG. The growth of the culture was monitored spectrophotometrically at a wavelength of 600 nm. Each graph is representative of at least two experiments.

FIG. 2.

Gef and ChpBK are bactericidal to E. coli. Cells of E. coli ERAPlacI harboring (A) pUPRIP, (B) pGef, and (C) pChpBK were grown overnight at 37°C and then diluted 1:100 into fresh LB medium containing appropriate antibiotics. At early log phase (OD₆₀₀ of 0.3 to 0.4), the culture was divided equally and incubated in the presence (▪) or absence (□) of 1 mM IPTG. At the indicated times (0, 2, and 4 h postinduction), samples from a dilution series of the culture were plated under repressed conditions on LB plates containing the appropriate antibiotics. Viable counts are expressed as mean CFU, counted on noninductive plates. Each graph is representative of at least two experiments.

FIG. 3.

In vitro phage delivery of lethal agents to E. coli. A CFU assay was performed to evaluate the effects of the phage-delivered lethal agents on the killing of E. coli ER2738. Target cells were grown to mid-exponential phase (OD₆₀₀ of 0.8) in LB broth containing tetracycline and diluted to approximately 10⁶ CFU/ml in LB broth containing 1 mM IPTG. An aliquot of cells (10⁵ CFU, 100 μl) was incubated at 37°C with an equal volume of phage lysate (8 × 10⁹ PFU/ml). Control experiments were carried out in the absence of phage lysate. Treatments were as follows: (first bar) cells plus buffer, (second bar) cells plus Gef phagemid lysate, and (third bar) cells plus ChpBK phagemid lysate. Treatments reflect viable cell counts following 30 min of incubation at 37°C. Viable counts were determined following dilution and plating of the infection on LB plates containing 1 mM IPTG. The figure is representative of at least two experiments with each infection performed in triplicate. All values are means ± standard deviation.

FIG. 4.

Cell death following phage delivery of pGef and pChpBK is IPTG dependent. A CFU assay was performed to evaluate the effects of the phage-delivered lethal agents in the presence and absence of IPTG. E. coli ERAPlacI cells were grown to mid-exponential phase (OD₆₀₀ of 0.8) in LB broth containing appropriate antibiotics (kanamycin and tetracycline) and diluted to approximately 10⁶ CFU/ml in LB with (▪) or without (□) 1 mM IPTG. An aliquot of cells (100 μl, 10⁵ CFU/ml) was incubated at 37°C with an equal volume of phage lysate (8 × 10⁹ PFU/ml). Control experiments were carried out in the absence of phage lysate. Treatments were as follows: (first bar) cells plus buffer, (second bar) cells plus Gef phagemid lysate, and (third bar) cells plus ChpBK phagemid lysate. Treatments reflect viable-cell counts following 30 min of incubation at 37°C. Viable counts were determined following dilution and plating of the infection on LB plates with (▪) or without (□) 1 mM IPTG. The figure is representative of at least two experiments with each infection performed in triplicate. The values are means ± standard deviation.

FIG. 5.

Phage-delivered lethal agents significantly reduce the level of viable bacteria in the blood of mice. A single dose of E. coli ER2738 (200 μl, 5 × 10⁸ CFU/ml), IPTG (100 μl, 250 mM), and phage preparation (200 μl, 1.2 × 10¹⁰ phagemid-containing particles/ml) was administered by intraperitoneal injection to cyclophosphamide-treated mice. Mice were treated with a single dose of phage preparation containing phagemid pGef, pChpBK, or pUPRIP (actual multiplicity of infection of 3.6). The control vector pUPRIP is identical to the phagemids pGef and pChpBK except that it lacks a gene encoding a lethal agent. At the indicated time points, blood samples were taken and bacterial counts were determined by plating onto LB plates containing tetracycline (20 μg/ml). Mice with tail blood containing less than 20 CFU/ml (lowest level of detection) at 1 h were eliminated from the analysis. The viable bacterial counts in the blood were plotted as the mean plus standard deviation for (A) each treatment group and (B) each animal within each treatment group. Statistical analysis with the unpaired t test indicated significantly fewer viable bacteria in the blood at the times indicated compared to the control pUPRIP group for the experimental pGef and pChpBK groups (*, P ≤ 0.05, two-tailed, adjusted for multiple comparisons).