In Vitro Activity of 3 Commercial Bacteriophage Cocktails Against Salmonella and Shigella spp. Isolates of Human Origin

Background: Salmonella and Shigella spp. are 2 of the most frequent and deadly enteric bacterial pathogens recorded worldwide. In developing countries Salmonella infections are responsible for many deaths annually and these mortality rates are prone to increase due to the emergence of resistance to antibiotics. In this overall scenario new alternative therapeutic approaches are needed. Methods: For the first time, we investigated the activity of 3 commercial bacteriophage cocktails (INTESTI, Septaphage, PYO) against a collection of contemporary Salmonella spp. (n = 30) and Shigella spp. (n = 20) strains isolated in Switzerland. Phage susceptibility was determined by implementing the spot test. Results: The overall susceptibility of Salmonella spp. to INTESTI and Septaphage was 87% and 77%, respectively. With regard to Shigella spp., the overall susceptibility to INTESTI and Septaphage was 95% and 55%, respectively. PYO was observed to be active against only 10% of Salmonella spp. but against 95% of Shigella spp. Conclusions: Our results seem promising, especially for the INTESTI biopreparation against Salmonella enterica infections. Nevertheless, such speculation should be supported by further in vivo studies to confirm efficacy and safety of the cocktails. We also emphasize the importance of large in vitro screening analyses aimed to assess the activity of such biopreparations against contemporary multidrug-resistant strains that are emerging worldwide.


INTRODUCTION
Salmonella and Shigella spp. are the most frequently found and deadly enteric bacterial pathogens. For instance, each year 500,000 cases of diarrheal shigellosis and about 1.2 million cases of nontyphoidal salmonellosis with 380 deaths are recorded in the United States [1][2][3][4]. Moreover, in developing countries Salmonella infections are responsible for 1 million deaths annually and these mortality rates are likely to increase due to the emergence of resistance to commonly implemented antibiotics [5,6]. In this overall scenario, new alternative and cost-effective therapeutic approaches are needed.
Bacteriophages are highly species-specific self-propagating viruses that can infect and lyse bacteria. Their employment is part of the standard medical practice in countries of the former Soviet Union, whereas in Western nations the use of phage therapy is unfamiliar, and this has led to a lack of studies analyzing efficacy and possible alternatives to antibiotics [7,8].
Numerous in vitro and in vivo reports exploring both lytic activity and clinical effectiveness to control Salmonella infections are available. However, such analyses have exclusively used monophages and focused on reducing contamination of food stuffs or intestinal colonization in food animals [9][10][11][12][13]. With regard to Shigella, Mai et al tested a phage cocktail (ShigActive TM ) in a mice model obtaining encouraging results [14].
To our knowledge, data regarding the in vitro activity of bacteriophage cocktails against large collections of Salmonella and Shigella spp. strains are still lacking. In this study, for the first time, we explored the in vitro activity of 3 commercially available bacteriophage cocktails currently implemented in the country of Georgia to treat human intestinal infections.
The collection of strains tested during the present study included contemporary Salmonella (n = 30) and Shigella spp. (n = 20) isolated from human infections which occurred in Switzerland. Species identification (ID) was routinely obtained using the matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS; Bruker). The ID confirmation and further typing were performed at the National Reference Laboratory for Enteropathogenic Bacteria and Listeria (Institute for Food Safety and Hygiene, Zurich, Switzerland). The antibiotic susceptibility profiles were obtained by disc-diffusion tests [17]. Most Salmonella spp. strains were pan-susceptible to tested antibiotics (ampicillin, ceftriaxone, cotrimoxazole, chloramphenicol, nalidixic acid, and ciprofloxacin), whereas only ceftriaxone was always active in vitro against isolates of Shigella spp. (Supplementary Table 1).
Phage susceptibility was determined with the spot test with double agar overlay method [18]. Briefly, 100 µl of a 0.5 McFarland bacterial suspension was mixed in a brain heart infusion (BHI) agarose matrix (0.6%), which was then distributed to solidify on a standard BHI agar plate. Then, 10 µl of each phage-suspension was spotted on the plate and incubated overnight. The day after, lysis zones were quantified [18]. Specifically, strains showing confluent lysis (complete clearing: ++++), semi-confluent lysis (clearing throughout, but with faint hazy background: +++), opaque lysis (turbidity throughout the cleared zone: ++), and taches vierges (individual clear or opaque plaques: +) were defined as susceptible to the phage compounds tested. Strains showing no activity (no clearing: R) were defined as resistant. For all strains (n = 50) susceptibility tests were performed in duplicate and on distinct days.

RESULTS AND DISCUSSION
As shown in Table 1, the overall susceptibility of Salmonella spp. to INTESTI and Septaphage was 86.7% (of which 23/30 were +++ or ++++) and 76.7% (none of which were +++ or ++++), respectively (examples in Supplementary Figure 1). With regard to Shigella spp., the overall susceptibility to INTESTI and Septaphage was 95% (of which 9/20 were +++ or ++++) and 55% (of which 3/20 were +++ or ++++), respectively. This data is promising, but we should note that the spot test can lead to an overestimation of the susceptibility as a consequence of the lysis-from-without phenomenon [19].
We did not expect any activity for PYO against our strains because, according to the manufacturer, this preparation should not contain lytic phages against Salmonella spp. and Shigella spp. However, we were surprised to note that this cocktail was active against 10% (of which 2/30 were +++ or ++++) of Salmonella spp. and, more importantly, against 95% (of which 7/20 were +++ or ++++) of Shigella spp. This could be explained by the presence of bacteriophages unable to selectively differentiate Salmonella and Shigella spp. from E. coli (all 3 being phylogenetically closely related bacterial species, especially the latter 2 [20]) that might share several common phage targets [21]. Moreover, taking into account the lysis-from-without phenomenon where a high multiplicity of infection can lead to bacterial death without infection, we are aware that by exclusively using the spot test, our susceptibility results might be slightly overestimated [19].  In conclusion, we showed the distinct spectrum and lytic activity of commercial bacteriophage cocktails targeting Salmonella and Shigella species. In particular, Septaphage proved to be active, though overall weakly, against 68% of the tested strains, whereas INTESTI exhibited a strong response against 90% of our isolates. Therefore, our results seem promising, especially for the latter biopreparation against Salmonella enterica infections. Nevertheless, such speculation should be supported by further animal studies together with human clinical trials in order to confirm efficacy and safety of cocktails. We also emphasize the importance of large in vitro screening analyses aimed to assess the activity of such biopreparations against contemporary multidrug-resistant strains emerging worldwide [2,22,23]. The sum of these steps, if successful, could lead to the maturation-also in Western countries-of an alternative approach for the treatment of bacillary dysenteries and salmonellosis. a Strains were defined as susceptible to the bacteriophages when confluent lysis (ie, complete clearing: ++++), semi-confluent lysis (ie, clearing throughout but with faint hazy background: +++), opaque lysis (ie, turbidity throughout the cleared zone: ++), taches vierges (ie, a few individual plaques: +) were recorded. Strains showing no activity (ie, no clearing "R") were defined as resistant.