Main text

Bacteriophages (phages) are obligatory bacterial parasites able to multiply exclusively in the presence of the host bacteria (Abedon 2009). The bacteriophage life cycle can be divided into several successive stages which are common to all known bacterial viruses. For lytic bacteriophages these include: adsorption (attachment), transfer of the genetic material from phage to host, expression and replication of phage genetic material inside infected cell, assembly of progeny (daughter) viral particles and their release (lysis of the host cell) to the environment and finally transmission to the new host (Clokie et al. 2011).

The accepted paradigm of the phage-host interaction states that the susceptibility of a bacterium to a particular bacteriophage is predominantly dependent on whether or not the bacteriophage can attach to specific attachment sites (so-called receptors) on the host (Bertozzi Silva et al. 2016). All bacteriophage infections are initiated through this irreversible binding process that involves a specific and tight contact between the virus particle and components present on the host cell surface. A fundamental understanding of the adsorption (attachment) mechanism is prerequisite to establish the nature of viral infection as well as the phage receptor-finding process is crucial for the efficient propagation of the infection and it determines its fate in environment (Marsh and Wellington 1994).

Despite the fact that phage-bacterial interactions have been studied for more than a century since the first discovery of bacterial viruses, still relatively little is known about the molecular basis of attachment (Miao and Miller 1999). Only a limited number of bacteriophage receptors have been identified, crystalized and characterized in detail. The majority of those well-characterized examples came from model systems including bacteriophages: λ and T4 infecting Escherichia coli and/or viruses infecting human pathogenic bacteria (e.g. Staphylococcus aureus and Listeria monocytogenes) (Bertozzi Silva et al. 2016). Conversely, in the case of the other viral-host systems, e.g. environmental and/or plant associated bacterial strains, this topic has been largely ignored so far (Okabe and Goto 1963; Rakhuba et al. 2010).

We propose here an inexpensive, high-throughput assay to preselect Pectobacterium spp. and Dickeya spp. genes involved in bacteriophage adsorption to host cells with the employment of a random transposon mutagenesis using a mini-Tn5 transposon combined with a screen for the active bacterial metabolism (visualized by reduction of resazurin to pink/yellowish-colored resorufin) in the presence of infectious bacteriophage particles.

Resazurin as an indicator of cell viability has been used so far in a number of applications including assessment of bacterial susceptibility to antimicrobials (Palomino et al. 2002), studies targeting microorganisms abundance in environment (Guerin et al. 2001) or measurements of aerobic respiration in ecosystems (González-Pinzón et al. 2012). Till present, the method has not been used however neither to assess survival of plant pathogenic bacteria in the presence of lytic bacteriophages, nor to search for Soft Rot Pectobacteriaceae (SRP) genes encoding bacterial proteins used by bacteriophages to interact with host bacterial cells during adsorption and infection.

SRP, including D. solani and P. parmentieri, are important necrotrophic plant pathogenic bacteria infecting a number of commercially relevant plant species worldwide (Pérombelon 2002; Charkowski 2007). In Europe, D. solani and P. parmentieri are recognized as emerging pathogens of important farming crops leading to increasing losses in agriculture (Toth et al. 2011; van Der Wolf et al. 2017). Likewise, SRP are recognized among the top-ten most important plant pathogenic bacterial species worldwide (Mansfield et al. 2012). For many years SRP have been a useful model for studying different molecular aspects of the plant-bacterium interactions for plants (Hugouvieux-Cotte-Pattat et al. 1996; Charkowski et al. 2012; Reverchon and Nasser 2013), these bacteria were however not used as a model to study viral- host interactions.

The aim of this study was to develop a fast and reliable screening technique to preselect candidate D. solani and P. parmentieri genes encoding bacterial proteins used by bacteriophages to interact with host bacterial cells during adsorption and infection.

To achieve this objective we selected as a model two type strains of emerging Soft Rot Pectobacteriaceae in Europe viz. D. solani strain IPO2222 (van Der Wolf et al. 2014) and P. parmentieri strain SCC3193 (Pirhonen et al. 1988; Khayi et al. 2016) and two characterized in detail in our previous studies lytic bacteriophages ϕD5 (Czajkowski et al. 2013; Czajkowski et al. 2014) – infecting IPO2222 and ϕA38 (Smolarska et al. 2017) – infecting SCC3193. Unless noted otherwise, well-known molecular biology and microbiological methods were used for all described experiments (Sambrook et al. 1989).

The methodology was based on a random transposon mutagenesis of IPO2222 and SCC3193 genomes with a mini-Tn5 transposon. To mutate bacterial genomes, conjugation of D. solani or P. parmentieri with Escherichia coli S17 λ-pir (transposon donor) carrying pFAJ1819 (Xi et al. 1999) suicide vector with mini-Tn5 was performed as previously described (Czajkowski et al. 2017). After conjugation, the resulting D. solani Tn5 and P. parmentieri Tn5 mutants were selected as previously described (Czajkowski et al. 2011) using M9 agar plates supplemented with 50 μg ml−1 neomycin (Sigma). All obtained IPO2222 and SCC3193 Tn5 transposon mutants were selected for assay with bacteriophages ϕD5 and ϕA38, respectively. For this, the bacteriophages ϕD5 and ϕA38 were propagated on their wild type bacterial hosts and titer as described earlier (Czajkowski et al. 2013). The adjusted stock concentration of ϕD5 and ϕA38 phage particles used in this study was 108 plaque forming units (pfu) ml−1 in tryptone soya broth (TSB; Oxoid). The individual D. solani and P. parmentieri Tn5 mutants and respective wild type strains (IPO2222 and SCC3193) were grown individually in 10 ml of TSB supplemented with 50 μg ml−1 of neomycin at 28 °C for 16 h with shaking (150 rpm) and after this time the bacterial cultures (containing ca. 108 colony forming units (cfu) ml−1) were diluted 100 times in the same fresh medium but not supplemented with the antibiotic. Such freshly diluted bacterial cultures contained ca. 106 cfu ml−1. Per Tn5 mutant to be analyzed, the wells of 96-well microtiter plate were inoculated with 50 μl of individual IPO2222 or SCC3193 Tn5 mutant culture (final bacterial density ca. 5 × 104 cfu per well) and subsequently were filled with 200 μl of bacteriophage solution (final phage particle concentration ca. 2 × 107 pfu per well). Each Tn5 mutant was analyzed in duplicates. To each well of the microtiter plate 5 μl of 0.22 μm-filter-sterilized 0.7% resazurin (Sigma) solution in demineralized water was added. As a control, 200 μl of sterile TSB instead of bacteriophage suspension was added to the wells inoculated with Tn5 mutants or wild type strains. Culture of a ɸD5 and ɸA38-resitant Escherichia coli strain DH5α (ThermoFisher Scientific) (Woodcock et al. 1989) grown in TSB at 28 °C for 24 h with shaking (150 rpm) and treated similarly as described above in case for Pectobacterium spp. and Dickeya spp. mutants and wild type strains, was used as an external assay control. After inoculation, 96-well plates were sealed with optically transparent sealing tape (Sarstedt) as suggested earlier (Lisicka et al. 2018) to prevent from contamination and evaporation of bacterial cultures during incubation. The plates were incubated for up to 12 h at 28 °C with shaking (150 rpm). After this time the development of the pink/yellowish color (positive reaction) in the inoculated wells was determined by eye (Fig. 1). The positive reaction indicates the growth of the Tn5 mutant in the presence of viable bacteriophages (which is equal to phage resistance) (Fig. 2). The experiment was performed twice with the same setup.

Fig. 1
figure 1

Example results of the fast screening assay to preselect Dickeya solani and Pectobacterium parmentieri Tn5 mutants in genes coding for proteins used by lytic bacteriophages ɸD5 and ɸA38 as receptors. Per Tn5 IPO2222 or SCC3193 mutant to be analyzed, in duplicates, the wells of 96-well microtiter plate were inoculated with 50 μl of individual IPO2222 or SCC3193 freshly-prepared Tn5 mutant culture in TSB (final bacterial density ca. 5 × 104 cfu per well) and subsequently with 200 μl of bacteriophage solution (final phage particle concentration ca. 2 × 107 pfu per well) and 5 μl of 0.7% resazurin (Sigma) solution in demineralized water. For control, per mutant, to the wells inoculated with bacterial Tn5 mutants or wild type strains as described above, 200 μl of the sterile TSB instead of bacteriophage suspension was added. The external assay control was culture of E. coli DH5α grow and propagated under the same conditions as described for IPO2222 and SCC3193. The pink/yellowish color indicates a positive reaction – bacterial growth in the presence of bacteriophage (phage resistance), whereas violet color indicates a negative reaction – no bacterial growth in the presence of bacteriophage (phage susceptibility)

Full size image
Fig. 2
figure 2

Host challenge assay of the parental (D. solani IPO2222 and P. parmentieri SCC3193) strains and selected Tn5 IPO2222 (A) and Tn SCC 3193 (B) mutants in the absence and presence of lytic bacteriophages ɸD5 and ɸA38, infecting D. solani IPO2222 and P. parmentieri SCC3193, respectively. Per mutant to be analyzed, in duplicates, the wells of the 96-wells plate were inoculated with 50 μl of freshly prepared bacterial cultures in TSB (final cfu ml−1 of 5 × 104). Subsequently, 200 μl of bacteriophage suspension in TSB was added per well (final pfu ml−1 of 2 × 107). Inoculated plates were incubated at 28 °C with shaking (150 rpm). For control instead of bacteriophage suspensions, the sterile TSB was added. E. coli strain DH5α, known to be resistant to lytic bacteriophages ɸD5 and ɸA38, was used as an external assay control. Growth rate was measured as a change in the optic density (OD) of the bacterial cultures at 600 nm wavelength for the total period of 12 h every 0.5 h using Epoch2 Microplate Spectrophotometer (BioTek Instruments)

Full size image

The candidate phage resistant Tn5 mutants selected in the experiments with resazurin were additionally screened using plaque formation assay and host challenge assay with the respective bacterial hosts (IPO2222 and SCC3193) (Clokie and Kropinski 2009) to confirm the phenotype of the each individual tested Tn5 mutant (Table 1).

Table 1 Plaque formation on lawns of D. solani IPO2222 Tn5 mutants and P. parmentieri SCC3193 Tn5 mutants
Full size table

The Tn5 mutants in genes coding for putative bacteriophage receptors and other proteins facilitating bacteriophage interaction with host cell were selected for sequencing the transposon insertion sites as previously described (Lisicka et al. 2018). In the preliminary screening, we selected the first five hundred D. solani Tn5 mutants and first 300 P. parmentieri Tn5 mutants for testing them for the resistance in the presence of lytic bacteriophages ϕD5 and ϕA38, respectively. From these, 3 (0.6%) and 3 (1%) of the tested D. solani and P. parmentieri Tn5 mutants, respectively, showed a bacteriophage resistant phenotype in assays with ɸD5 and ɸA38. Example results are shown in Figs. 1 and 2 and Table 1.

The proposed fast screening of bacteriophage-resistant bacterial mutants allows direct pre-selection of candidate mutants for the follow-up studies. These studies may include the analyses of interactions of the phage-resistant D. solani and/or P. parmentieri Tn5 mutant with host plants and the impact of various biotic and abiotic stresses and environmental conditions on phage-resistance bacterial mutants.

The main advantage of the suggested method is that the assay can be considered as a definitely fast preliminary screening (the total assay time is ca. 12 h) for the identification of the genes coding for bacterial proteins used by bacteriophage to interact with host cells. What is more, such Tn5 mutants may be subsequently tested against several other factors of interest. The method does not require any expensive laboratory tools and consumables and therefore it may be performed in virtually every microbiological laboratory possessing standard microbiological equipment.

Likewise, it has to be noted here that the proposed fast screening assay may be used not only to evaluate the resistance/susceptibility of a particular Tn5 bacterial mutant for phage infection but also in ecological studies to screen a number of environmental bacterial isolates against a collection of different viruses to assess the host range of a particular phage of interest.

We postulate that due to the universal nature of this screening method, it can be easily adapted to study interaction of other bacterial species and other lytic bacteriophages of interest.