Using experimental evolution to create phage therapy agents to target the Horse Chestnut bleeding canker pathogen, Pseudomonas syringae pv. aesculi

A bleeding canker disease epidemic is afflicting the circa 470,000 ecologically and socially important European Horse Chestnut (Aesculus hippocastanum) trees in Britain. The disease is caused by the bacterial pathogen, Pseudomonas syringae pv. aesculi (Pae), which infects phloem and xylem vessels in the trunk and branches, causing 'bleeding' as tree exudates flow from infected sites (Green et al. 2009 Plant Pathology 58, 731-744); it can eventually lead to tree death thus posing the potential for another 'Dutch Elm' disaster. Bacteriophages have enormous potential for treating bacterial disease because of the large diversity and the ability to evolve multiple genotypes in the laboratory to use as multi-genotype mixtures. Phage therapies could also be used as a prophylactic application to act as a further barrier to infection. We propose to sample phages to: analyse diversity; evolve phage with its host to create new genotypes; analyse the efficacy of phage in treating trees. Objective 1: Sampling and diversity To sample environmental Pae phages, phages will be isolated from healthy and diseased tree tissue and from the surrounding soil and plated onto Pae lawns in semi-solid overlays on agar plates: lytic phages will be identified by their plaque morphology. Plate counts will determine the abundance of Pae-infecting lytic phages. Lytic phages will be characterised and compared using microscopy and molecular techniques. Objective 2: Experimental evolution (two stages) Stage 1 - Perform a large screen to assess the coevolutionary potential of isolated phages. Three replicate populations of each unique phage isolate (if >50 a representative subsample of isolates will be used) and Pae will be propagated in microcosms (30ml glass universals containing 6ml of KB liquid medium) by batch culture (1% of each population will be transferred to a fresh microcosm every 48 hours) for 12 transfers. Every 2nd transfer population samples will be frozen at -80C in 20% glycerol, and phage samples will be isolated by treatment with chloroform and stored at 4C. Every 4th transfer we will measure coevolution using a standard time-shift protocol (Brockhurst 2003 Ecol Lett 6:975-979). Bacteria-phage coevolution is characterised by the evolution of increasingly broad infectivity and resistance ranges through time. Therefore, significant coevolution will be defined if in replicate populations and for multiple timesteps there is a positive slope of phage infectivity and bacterial resistance with time. The 5 phages that display the greatest coevolution will be selected for further experiments. Stage 2 - Perform two types of long-term selection experiment: [1] coevolution, which selects for increased phage host range (i.e. phage that can infect a wide range of bacterial genotypes), experiments will be performed as described above for 50 transfers, standard time shift experiments will be used to assess the progress of coevolution; [2] passaging on a fixed bacterial genotype, which selects for high rates of host exploitation, experiments will be performed by batch culture whereby at each transfer only the phage is transferred from the previous to the new microcosm, and the bacteria are replaced by the ancestral genotype grown up from the laboratory stock. After 50 transfers the response to selection will be assessed in evolved phage lines by performing one-step growth curve experiments to determine key life history parameters (i.e. adsorbtion rate, lysis time, burst size). Objective 3: Therapy trials The efficacy of phage therapies in trees will be done using trees grown in controlled environment growth rooms supplied by the CASE partner, BTE. Different titers of single and mixed phage inocula will be applied to infected and pre-infected trees using different modes of application (spray and high pressure injection) to determine the efficiency of phage as a prophylactic as well as a therapy.