Bacteriophage-based biocontrol treatments of xanthomonads intend to interfere with a plant-pathogenic Xanthomonas spp. system. This subsection contains essential information on this system.

Xanthomonads live part of their life cycle outside the host plant as epiphytes in the lesions of fallen leaves or associated to plant debris in the soil [51]. They are an essential component of the soil microbiome, with 2–7% relative abundance in the bacterial community [52].

The infection cycle of Xanthomonas spp. starts with an epiphytic phase followed by entering the host plant through natural openings (stomata, hydathodes) and wounds to start its internal colonization (endophytic phase) [53]. When introduced into the plant surface, xanthomonads use a variety of adhesion strategies to attach to the plant ^{[54][55][56][57][58][59]}[54,55,56,57,58,59]. Plants have also evolved various defence mechanisms to protect themselves from pathogens [60]. They respond to pathogen associated molecular patterns (PAMPs) by activating PAMP-triggered immunity (PTI) or effector-triggered immunity (ETI) mediated by pathogen-specific receptors [61]. As a result, a systemic acquired resistance (SAR) status may be established, potentially increasing resistance to subsequent attacks in the entire plant ^[62][63][64][62,63,64].

A first key element of bacterial survival in the phyllosphere is the biofilm formation, creating a microenvironment that can protect bacteria against environmental stress conditions ^[58][65][58,65]. This is an important virulence factor of phytopathogenic Xanthomonas spp. ^[66][67][66,67]. A biofilm, in addition to the cells, is primarily made up of proteins, lipids and extracellular polysaccharides (EPS) ^[68][69][68,69]. The formation of a biofilm may provide resistance to host defence mechanisms and vascular bacteria attachment to xylem vessels, or contribute to bacterial epiphytic survival prior to colonisation of the plant intercellular space [70]. The gum operon, a massive transcriptional unit containing 12 enzyme coding genes (gumB-gumM), mediates xanthan gum biosynthesis [71]. A study revealed that biofilm production deficient mutants (particularly gumB and gumD) showed significantly lower leaf surface survival than wild type X. citri pv. citri and X. axonopodis pv. manihotis ^[72][73][74][72,73,74]. The study of many Xanthomonas spp. have shown that gum genes contribute to bacterial in planta growth, epiphytic survival and disease symptom formation ^{[72][75][76][77][78]}[72,75,76,77,78].

The assembly and dispersal of biofilms are partly mediated by the Quorum-sensing (QS) signal molecule, or diffusible signal factor (DSF). DSF positively influences the disruption of biofilms [79].

One survival strategy of bacteria during unfavorable conditions is the formation of persister cells. Persisters are a small fraction (0.001%–0.1%, or up to 1% in biofilms) of cells in a metabolically inactive, dormant state that are resistant against a wide range of antibiotics [80]. X. campestris pv. campestris and X. citri subsp. citri can form persister cells under different stress conditions ^[81][82][81,82]. Importantly, bacteriophages can also infect persisters [83].

LPS, as major components of the bacterial outer membrane, protect the cell from harmful environments and are another surface-associated virulence factor in Xanthomonas spp. Importantly, LPS not only function as virulence factors but also induce plant defense responses, such as pathogenesis-related gene expression, cell wall thickening and oxidative burst ^[84][85][84,85]. Mutations in LPS gene clusters make bacteria more susceptible to adverse environmental conditions, which may result in a reduction in bacterial virulence, as shown for X. campestris pv. campestris ^[86][87][88][86,87,88].

Xanthomonas species have a plethora of potential mechanisms that aid bacterial fitness in diverse environments, including the six different extracellular protein secretion systems (referred to as type I–VI, or T1SS–T6SS) that export proteins via the bacterial multilayer cell envelope and, in some cases, into host target cells. The conserved structural components that characterize these secretion systems, as well as the characteristics of their substrates and the pathway that these substrates take during the export process, distinguish them. T6SS was recently discovered and is involved in at least 25% of all sequenced gram-negative bacterial genomes [89]. The Hcp and VgrG proteins are essential components of T6SS that mimic the bacteriophage tail and needle complex, respectively [90]. Yang et al. [90] investigated the evolution of the T6SS in the Xanthomonas genus and assessed the relevance of the T6SS for virulence and in vitro motility in X. phaseoli pv. manihotis (Xpm), the causal agent of cassava bacterial blight. According to their phylogenetic analyses, the T6SS may have been obtained through a very ancient event of horizontal gene transfer (HGT) and preserved through evolution, implying their significance for host adaptation. They also showed that the T6SS of Xpm is functional and immensely contributes to motility and virulence.

Transcription activation-like effectors (TALEs) ensure plasticity in host adaptation for xanthomonads. TALEs have a repetitive domain governing the binding to promoters of host genes [91]. Novel TALEs could be created because this repetitive region is shared among TALEs, and recombination frequently occurs, as it was recently demonstrated in X. oryzae pv. oryzae [92]. These novel TALE encoding genes could be changed by HGT between bacteria, strengthening their host adaptation abilities [93].

When investigating ecological roles of bacteriophages in a Xanthomonas spp. population, it should be highlighted that the relationship between bacteriophages and their hosts could be both antagonistic and mutualistic, and the long-term survival of a bacteriophage population does not always require the lysis of its host. Therefore, bacteriophages are not predators, but either parasites or parasitoids of the host [94].

Bacteriophages can infect bacteria located in biofilms, albeit biofilms can provide a barrier for bacteriophage attacks compared to planktonic bacteria. This barrier is due to the physiological heterogeneity of the bacteria composing the biofilms, the secreted EPS, and the differential display of receptors on the host cell’ surface [94]. Bacteriophages can interact with biofilms of xanthomonads at several points. In a recent study Yoshikawa et al. [37] isolated the X. citri jumbo bacteriophage XacN1. They showed that the XacN1 genome encodes potential lytic enzymes such as cell wall hydrolases, C1 family peptidase, M23 family peptidases, lipase and chitinase. According to proteomic analysis, lipase, chitinase, and M23 family peptidases were discovered in the bacteriophage XacN1. They concluded that these enzymes may be necessary to disrupting the biofilm and initiating bacteriophage infection. Bacteriophages have evolved to counteract the biofilm barrier by using depolymerase enzymes on their capsids, and can also induce host lysis, allowing bacteriophages to degrade biofilm [95]. Furthermore, bacteriophage genomes carrying QS genes were detected in Clostridium difficile bacteriophage phiCDHM1 and three Paenibacillus bacteriophage genomes ^[96][97][98][96,97,98]. These genes can modify the biofilm disruption and other QS-mediated responses, including the decision on the lysogenic or lytic lifecycle of bacteriophages [97] or even the synthesis of virulence genes, as demonstrated in X. campestris [98].

Generally, the diversity of bacterial communities can support their adaptation to environmental circumstances [99]. If a community is more diverse, it is more stable as it can better adapt to the changing environment [100]. Prokaryotic viruses are essential in driving processes in microbial ecosystems ^[101][102][101,102]. In the absence of bacteriophages, one or several strains could become dominant in the niche, and other strains could be extinct, as was demonstrated in in vitro experiments ^[101][103][101,103]. Bacteriophages most likely infect the most abundant host strain, causing a decrease in its abundance (”kill the winner” principle). A consequence of this action will be a fluctuating selection, that increases diversity [103] and strengthens the community’s stability or adaptation ability. This may cause that bacteriophage-based pesticides can support the presence of xanthomonads on the fields when not applied carefully. Integrated disease management together with the application of carefully selected bacteriophages timed appropriately could be one solution.

The genome of lysed cells will be available for surrounding bacteria, providing them novel genetic information, which may also include pathogenicity-related genes, as recently shown in the case of the cherry pathogen Pseudomonas syringae pv. morsprunorum or in X. albilineans ^[104][105][104,105]. Lytic bacteriophages increase the mutation rate in their host’s genome, even in genes not related to bacteriophage resistance/immunity [101]. This effect can drive both adaptation (short term) or evolution (long term) processes. These from point of biocontrol disadvantageous features of lytic bacteriophages (i.e., providing novel genetic material for surrounding bacteria, increasing the mutation rate in the host’s genome) could be managed by an integrated disease management. However, the mentioned drawbacks are less serious, for example, when lysogenic bacteriophages are applied in the fields. Lysogenic bacteriophages can protect bacteria carrying their genomes from superinfection (Superinfection: A second (delayed) bacteriophage infection of an already bacteriophage-infected bacterium) [106]. Horizontal gene transfer is one of the major factors (together with the mutations in avirulence genes) to evade host resistance ^{[107][108][109]}[107,108,109]. The fact that 5–25% of the genome of Xanthomonas spp. originates from recombination events [110] highlights its importance in xanthomonads evolution and adaptation processes. Exchange of virulence factors between Xanthomonas spp. via HGT was observed in several cases [12].

The complexity of these HGT actions is demonstrated in the genome of a X. anoxopodis strain that contains a truncated bacteriophage genome carrying a gene resembling a plant protein that is induced during citrus blight disease [111].

As bacteriophages are often strain-specific, they can also act on the population level, influencing the population’s intraspecific composition. Consequently, lysogens can contribute to the colonization of new niches. When lysis is induced in a small portion of the lysogenic cells, from superinfection-protected bacterial populations, and the bacteria originally located in the niche to be colonized are not protected from the infection, the new population can use their lysogenic bacteriophages as a weapon against the indigenous cells (“kill the relatives” principle) [101]. On the contrary, native bacteria can protect themselves against colonization by sacrificing a part of the population and inducing their prophages’ lytic cycle [101]. Lysogenic bacteria can use their prophage weapon effectively, as observed in an in vitro experiment recently, where a lysogenic-lytic switch of bacteriophages to QS autoinducers strongly influenced the viral and bacterial abundance and diversity in soil communities [112].

There are examples of how lytic induction is carried out to optimize the multiplicity of infection (MOI). QS, encoded by either bacteria or bacteriophages, can influence this process ^[113][114][113,114]. Moreover, some bacteriophage genomes contain their own density monitoring equipment (the arbitrium system) and encode for small oligopeptides with which the bacteriophage density can be measured, as described in Bacillus bacteriophages ^[114][115][114,115]. Lysogeny is preferred when bacteriophages are abundant. Based on the described features of lysogenic and transducing bacteriophages, their field application may contribute to the adaptation and pathogenicity of xanthomonads, i.e., it may lead to unwanted effects. Therefore, the application of well-characterized, strictly lytic bacteriophages is advisable for bacteriophage-based biocontrol.

As bacteriophages and their hosts are not alone in the microflora, bacteriophages will meet their hosts with rare frequency when the living cell number of the host is low. Thus, one important consequence of the ”kill the winner” principle is that bacteriophages cannot reduce the living cell number of their hosts to zero in a community [116], a property which differs from most chemical antibacterial compounds.

We mentioned examples in this subsection, how bacteriophages (both lytic and lysogenic ones) can alter the strain and/or species abundancies in communities. The composition of Xanthomonas spp. population and/or the microbial community may be distinct in different fields which may be differentially influenced by the described effects of bacteriophages. In addition to the environmental factors, a result of this divergent influence may lead to a distinct outcome of bacteriophage-based biocontrol in fields, at least in several cases [117].

Bacteriophage-resistance mutations in bacteria usually come with a fitness cost, such as a decrease in virulence, which results in less disease severity. This is because many of the molecules taking part in bacteriophage attachment are also engaged in the virulence mechanism. As a result, mutations that lead to resistance commonly compromise virulence. There are a few examples of how mutations in bacteria surface structures lead to decreased virulence, such as mutation in the X. campestris xanA gene needed for xanthan and lipopolysaccharide synthesis, which significantly decreases the effectiveness of bacteriophage L7 adsorption [118].

Bacteriophage resistance in bacteria is one of the main concerns regarding the bacteriophage-based biocontrol strategies. A detailed understanding of bacterial resistance to bacteriophages and their interaction with plants play an important role in the design of bacteriophage-based biocontrol strategies of xanthomonads. To survive bacteriophage infections, bacteria have developed a wide range of protection strategies, including spontaneous mutations, restriction modification systems (R–M systems), and adaptive immunity through the CRISPR-Cas system [106]. The key mechanisms driving bacteriophage resistance are spontaneous mutations, which can grant bacteriophage resistance by altering the structure of bacterial surface components that function as bacteriophage receptors [119]. Furthermore, bacteria can acquire resistance through lysogenic bacteriophages that carry sequences in their genetic material which encode bacterial resistance or toxins and incorporated into the bacterial genome [120]. The mechanisms by which bacteriophages counteract the anti-bacteriophage systems of bacteria are poorly understood. Bacteriophages with the ability to acquire new receptor tropism can modify their receptor-binding protein, which means that when a host receptor changes to a mutated form, bacteriophages can recognize the altered receptor structure and thus overcome disturbance in receptors for bacteriophage adsorption [121]. Bacteriophages use various anti-restriction strategies to avoid the wide range of R–M systems. These modification genes encode a small protein that is transmitted to the cell with the viral genome, or it may instantly neutralize the host immune system by intervening with the formation or function of the CRISPR–Cas ribonucleoprotein [122]. Bacteriophages may use bacterial CRISPR–Cas systems to promote their own replication, allowing the phage to complete its lytic cycle [123]. When a bacterium develops resistance to a specific bacteriophage, it retains sensitivity to bacteriophages with various cell surface receptors. Bacteriophage-mediated selection can be used in disease management, for example, by combining various bacteriophages to broaden the host range and suppress resistance evolution [124] and/or reasonably combining bacteriophages and chemical control to establish synergies and decrease the likelihood of resistance evolution [125]. This implies that the application of a bacteriophage cocktail may be beneficial, even if bacteria quickly develop resistance, since resistant strains may be less fit, thus more treatable using another combined method.

3. Bacteriophage-Based Biocontrol of Xanthomonas spp.

3.1. Examples for Greenhouse and Field Trials

Shortly after their discovery, bacteriophages were evaluated for control of plant diseases, including those caused by Xanthomonas spp. Some of the first studies were conducted by Mallman and Hemstreet (1924) who isolated the “cabbage-rot organism” X. campestris pv. campestris from rotting cabbage and showed that the filtrate from the decomposed tissue could inhibit pathogen growth in vitro [126].

From the 1960s, a considerable number of studies explored the efficacy of phages for the control of bacterial spot of peach, caused by X. arboricola pv. pruni ^{[127][128][129][130]}[127,128,129,130]. Civerolo and Keil [127] applied bacteriophages 1 h prior to inoculation by the pathogen and reduced bacterial spot severity on peach leaves to 22% compared to 58% for control plants under greenhouse conditions. Civerolo [128] found that preinoculation of peach seedling foliage with crude lysates of the bacteriophage mixtures resulted in 6–8% fewer infected leaves and a 17–31% reduction of disease compared to control plants. Application of premixed bacteriophage—pathogen suspension immediately before inoculation resulted in a 51–54% decrease of bacterial spot symptoms in peach seedlings. Zaccardelli et al., isolated eight bacteriophages active against X. arboricola pv. pruni, examined their host range and lytic ability, and selected a lytic bacteriophage strain with the broadest host range for disease control ^[129][130][129,130]. By weekly bacteriophage treatment they significantly reduced fruit spot incidence on peaches [130].

Significant achievements have been made in bacteriophage application for control of bacterial spot of tomato caused by X. campestris pv. vesicatoria in greenhouse and field conditions ^{[131][132][133][134][135][136][137][138]}[131,132,133,134,135,136,137,138]. Flaherty et al. [131] used a mixture of host range mutant bacteriophages and effectively controlled tomato bacterial spot in greenhouse and field conditions. Moreover, bacteriophage application increased total weight of extra-large fruit comparing to nontreated control or plants treated with chemical bactericides. Balogh et al. [133] improved the efficacy of bacteriophage treatments in field and greenhouse experiments by using protective formulations that significantly increased bacteriophage longevity on the plant surface. Bacteriophage mixture formulated either with 0.5% pregelatinized corn flour, Casecrete NH-400 with 0.25% pregelatinized corn flour, or 0.75% powdered skim milk with 0.5% sucrose, provided significant disease control compared to untreated control. However, in greenhouse experiments skim milk gave the best results, while Casecrete performed best in the field [133].

In order to improve bacteriophage efficacy and provide consistent disease control, bacteriophages of X. campestris pv. vesicatoria have been studied as a part of integrated disease management practices [138]. Obradovic et al., tested various combinations of plant inducers and biological agents for control of tomato bacterial spot [139]. Acibenzolar-S-methyl applied in combination with bacteriophages formulated with skim milk and sucrose, reduced bacterial spot of tomato in a greenhouse [136] as well as in the field [135]. Recently, Abrahamian et al. [140] evaluated 19 different chemical agents, biological control agents, plant defense activators, and novel products for their ability to manage bacterial spot on tomato caused by X. perforans. They reported that combination of bacteriophages, cymoxanil, famoxadone and phosphoric acid, significantly improved the disease management compared to the copper-based standard treatment. All these studies led to bacteriophage treatment, integrated with other disease management practices (e.g., late blight), becoming a part of a standard integrated management program for tomato bacterial spot in Florida ^[138][139][138,139].

Gašić et al. [141] studied the efficacy of bacteriophage KФ1 in the control of pepper bacterial spot caused by X. euvesicatoria. They found that double bacteriophage application, before and after challenge inoculation, significantly reduced disease incidence when compared to untreated control. However, integrated application of bacteriophages 2 h before and copper hydroxide 24 h before inoculation was the most efficient treatment. The same bacteriophage strain was used as a part of integrated disease management and combined with other biocontrol agents, copper compounds, antibiotics and plant inducers to control pepper bacterial spot [142]. Bacteriophage combination with copper-hydroxide and acibenzolar-S-methyl was the most effective treatment reducing the disease severity by 96–98% compared to control [142].

Similar studies were performed to develop management strategies for efficient and sustainable control of leaf blight of onion, caused by X. axonopodis pv. allii. Lang et al. [143] reported that biweekly or weekly applications of bacteriophages reduced disease severity in the field by 26 to 50%: similar to results achieved by weekly applications of copper-mancozeb. Therefore, integrated application of bacteriophage mixtures with acibenzolar-S-methyl could be a promising strategy for managing Xanthomonas leaf blight of onion and contribute to reduced use of chemical bactericides [143].

Comprehensive research was done on bacteriophage-mediated control of Asiatic citrus canker caused by X. axonopodis pv. citri, and citrus bacterial spot X. axonopodis pv. citrumelo ^{[144][145][146]}[144,145,146]. Bacteriophage treatment, without skim milk formulation, provided an average 59% reduction in citrus canker severity in greenhouse experiments. In nursery, bacteriophage treatment reduced disease, but was less effective than copper-mancozeb, while bacteriophage integration with copper-mancozeb resulted in equal or less control than copper-mancozeb application alone [145]. Similar results were obtained in the management of citrus bacterial spot, where bacteriophage treatment provided significant disease reduction on moderately sensitive Valencia oranges while it was ineffective on the highly susceptible grapefruit [145]. Ibrahim et al. [146] reported that successful control of Asiatic citrus canker in greenhouse and field can be obtained by combination of bacteriophage mixture formulated with skim milk-sucrose and acibenzolar-S-methyl.

Initial research of bacteriophage infecting X. oryzae pv. oryzae, the causal agent of bacterial blight of rice, was conducted by Kuo et al., who applied purified bacteriophages 1, 3, and 7 days before inoculation, and obtained 100%, 96% and 86% reductions of bacterial leaf blight, respectively [147]. Recently, Chae et al. [148] significantly reduced the occurrence of bacterial leaf blight to 18.1% compared to 87% in untreated control by treatment with skim milk formulated bacteriophages. Ogunyemi et al. [149] reported the bacteriophage X3 was more effective in disease severity reduction (83.1%) if sprayed before inoculation rather than after (28.9–73.9%) it. However, seed treatment with bacteriophages reduced disease by 95.4%.

Other results on using bacteriophages specific to Xanthomonadaceae in plant disease control includes reduction of incidence of bacterial blight of geraniums caused by X. campestris pv. pelargonii with foliar application of h-mutant bacteriophages [150]. Nagai et al. [151] found that a non-pathogenic Xanthomonas sp. strain mixed with bacteriophages effectively controlled black rot of broccoli caused by X. campestris pv. campestris in field trials. Orynbayev et al. (2020) studied effects of bacteriophage suspensions mixed with different UV-protectants in control of black rot caused by X. campestris pv. campestris on cabbage seedlings. In two-year greenhouse experiments, bacteriophage DB1 mixed with 0.75% skimmed milk showed an average efficacy of 71.1% in control of the disease, compared to 59.1% efficacy of Kocide 2000 treatment [152].