To date, most biocontrol studies have used phages for the decontamination of food and limited data are available concerning the stability of phages in an environmental matrix. Furthermore, the use of a phage to prevent infections caused by MDRAB has not been demonstrated. The ϕAB2 phage was selected as a model phage for this study because its DNA and protein profiles were previously determined . The current study demonstrated that phages such as the ϕAB2 phage might be useful for reducing MDRAB contamination in liquid suspensions or on hard surfaces such as may be encountered in ICUs, and may be added to a solution to produce an antiseptic hand wash.
One issue with the human use of phages is their potential toxicity. Previously, we demonstrated ϕAB2 had 91–99% DNA sequence identity with the fully sequenced ϕAB1 and that to date, no putative or confirmed toxin genes have been identified in ϕAB2 . In addition, no prophage-related genes were observed in ϕAB1, although Vallenet et al. suggested that putative prophage sequences account for 5.1% and 6.7% of the genomes of both A. baumannii strains . Thus, it is reasonable to assume that ϕAB2 has no toxin genes or prophage-related genes, and we predict there will no safety issues related to toxin production or chromosomal integration of ϕAB2.
There have been limited studies regarding environmental effects on phage stability. A previous study investigated another A. baumannii-specific phage, AB1, which is relatively heat resistant and can survive temperatures of 50–60°C, and even a 15-min incubation at 90°C . The stability of ϕAB2 at extremely high temperatures was not evaluated in the present study because ϕAB2 is proposed for use as an alternative sanitizer, so information regarding its stability for long storage periods at refrigerated or freezing temperatures was more relevant. Our study demonstrated that phage infectivity is strongly dependent on environmental conditions such as temperature, pH, and the presence of other organic substances. Investigation of the optimal pH for maintaining ϕAB2 infectivity demonstrated that the least damaging pH tested was pH 7, similar to the sewage from which ϕAB2 was isolated (pH 7.8). Yang et al. also demonstrated that the AB1 phage was most stable at pH 6, and that less than 42.9% of AB1 phages lost their infectivity in a range between pH 5–9 . Thus, our environmental stability results indicated that ϕAB2 should be stored in a pH 7 solution at 4°C for extended storage if ϕAB2 is incorporated into a detergent.
Because MDRAB survives for long periods on environmental surfaces and may promote cross-transmission, we investigated the efficiency of ϕAB2 in reducing A. baumannii M3237 contamination on surfaces. We observed the ϕAB2 concentration required to reduce A. baumannii M3237 contamination was lower for liquid suspensions than hard surfaces. The mean survival rate ratio of A. baumannii M3237 between surface and liquid suspension ranged from 2–10,151 depending on the phage concentration. As ϕAB2 does not diffuse as freely on a hard surface as in a suspension, a higher concentration of ϕAB2 was required for surface decontamination of MDRAB compared with in solution.
The ability of phages to persist on a surface for extended periods is limited by many factors, such as desiccation , which may explain the loss of ϕAB2 infectivity after 2 months storage on a glass surface. Because ϕAB2 cannot survive for long periods on a hard surface, the phage detergent must be frequently re-applied to surfaces to provide persistent bactericidal or MDRAB activity.
Previous biocontrol studies suggested that high phage numbers should be used without relying on phage amplification [22, 23]. Although ϕAB2 has a larger burst size than other phages [23, 35], it is important to determine the optimal phage concentration that will allow efficient phage attachment and amplification for the quantity of MDRAB present. Experiments on environmental ICU samples have identified A. baumannii on 39% of the sampled surfaces with a mean A. baumannii DNA concentration of 19,696 copies . Based on the results of our surface evaluation, we recommend that a phage concentration of at least 107 PFU/cm2 be applied to surfaces in ICUs. This approach may not be suitable for the treatment of large surfaces, but may be useful for small biomedical devices. Abuladze et al. suggested a glass matrix is easier to decontaminate than gypsum . Thus, the phage decontamination efficiency for different surfaces such as gypsum, plastic, Teflon, or other polymers may vary, and requires further investigation. In addition to phage concentration, the incubation time is also critical for surface applications. When a high phage concentration (108 PFU/slide) was used to treat a surface contaminated with bacteria at a concentration of 105 CFU/slide, an incubation time of 5 min resulted in a 96% reduction of A. baumannii M3237 numbers. This incubation time was caused a 94% reduction in the number of Escherichia coli O157:H7  under the same test conditions.
MDRAB can be transmitted via the hands of health-care personnel. However, frequent or improper hand washing can cause skin to lose moisture or become irritated, reducing the hand washing rate despite intensive hand washing educational programs. Therefore, the addition of paraffin oil and glycerol to the formulation of antiseptic hand wash might retain skin moisture and reduce the transmission of MDRAB from the hands of health-care workers to patients. Alcohol-based hand rubs could reduce skin irritation  and reduce the number of bacteria more effectively than soap and water in a number of experimental models [42, 43]. However, A. baumannii may metabolize low levels of alcohol to become more virulent . Thus, an alternative hand washing approach is required to prevent microorganisms becoming tolerant to alcohol-based disinfectants in the future.
In this study, we designed two antiseptic hand wash experiments and observed a difference in the bactericidal effect between phage-containing lotion and glycerol solution, possibly related to the stability of ϕAB2 in different media. Because the detailed compositions of commercial creams are proprietary, it is difficult to explain the unpredictable changes of phage numbers in the cream, as phages could aggregate, disaggregate, or decay after long storage periods. O’Flaherty et al. demonstrated that S. aureus-specific phage K exhibited antibacterial activity when incorporated into a bismuth-based cream . The bismuth cream exhibited well antibacterial activity, but the related phage stability was not reported. In contrast, we observed that ϕAB2 was stable in 10% glycerol after 90 days storage at room temperature. Glycerol is a common cryoprotectant for phage infectivity during storage at temperatures between −20 and −70°C. Other phages, including F-specific RNA bacteriophages, and Bacteroides fragilis-specific phages, are also stable in 10% glycerol for up to 50 days  and can retain their infectivity with even longer storage times.