The honey bee (Apis mellifera) is an important element of natural environment that play a vital role in the process of pollination, and contributes to the improvement and maintenance of flora biodiversity (Morse and Calderone, 2000). Bees’ activity also provides such valuable products as honey, bee pollen, propolis, bee wax, and royal jelly, which are widely used by humans in various industries, including food and diet supplement production, cosmetology, natural medicine and pharmacology. Unfortunately, a significant decrease in the number of bee colonies has been observed worldwide in the last few decades. Much attention has been given to colony collapse disorder (CCD), described as an abnormal phenomenon based on the disappearance of the majority of worker bees in a colony; only the queen, lots of food and a few nurse bees remain in the nest to care for the remaining immature bees (van Engelsdorp et al., 2009; van Engelsdorp and Meixner, 2010). In recent years, honey bees have been exposed in the environment to many adverse factors, which include the chemicalization of modern agriculture, large-scale use of plant protection products, environmental degradation, as well as diseases caused by different pathogens and parasites (Cox-Foster et al., 2007; Dainat et al., 2012). In reality, all of these factors tend to overlap and interact, which means that their synergistic action can cause health problems in bee colonies, such as the abrupt disappearance of worker bees from the colony.

American foulbrood (AFB), caused by Paenibacillus larvae, is one of the most infectious, dangerous, lethal and easily spreading diseases of Apis mellifera caused by different pathogens and parasites. Despite the name, AFB is classified as a notifiable disease with a worldwide distribution in almost all beekeeping regions in each of the five continents (Alippi and Aguilar, 1998). The causative pathogen has been described by White (1906) as Bacillus larvae, a Gram-positive, spore-forming bacterium (Genersch et al., 2006) that can produce even more than one billion spores per infected larva (Shimanuki, 1997; OIE (World Organization for Animal Health), 2018). Infectious spores are transferred within or between colonies by worker bees or by beekeeping practices (Sturtevant, 1932; Lindström et al., 2008a, b). Spores are extremely long-lived (they can survive even more than several decades in honey or on hive equipment (Hasemann, 1961; Genersch, 2010) and resistant to unfavorable conditions, e.g., heat and chemical agents, thus they are very hard to remove (Genersch, 2010). Unfortunately, conventional antimicrobial therapies are only effective for the vegetative forms of bacteria, and so far AFB has proved impossible to eradicate anywhere using all available and allowed methods of treatment and prevention. The use of antibiotics to treat AFB is not a permanent solution due to the production of resistant spores and increase in antibiotic resistance in bacterial cells (Lodesani and Costa, 2005; Alippi et al., 2007); they can also contaminate honey, which could be dangerous for humans consuming this product (Ortelli et al., 2004; Martel et al., 2006; Saridaki-Papakonstadinou et al., 2006; Meeraus et al., 2015; Muriano et al., 2015). This is the reason why antibiotic application in AFB treatment has been banned in most European countries (Genersch, 2010; Forsgren et al., 2018). In many countries, disease control even includes burning of infected colonies that generates huge economic losses. Problems associated with the control and treatment of infected colonies result in a significant decrease in honey bee populations, beekeeping industry and, in consequence, agricultural production all over the world.

Bacteriophages may be a promising solution in the treatment and prevention of AFB spread in honey bees (Tsourkas, 2020). These bacterial viruses are commonly found in the biosphere (Clokie et al., 2011). A recent study has demonstrated that phages (both lytic and temperate) may be part of the honey bee gut microbial community (Bonilla-Rosso et al., 2020), participating in its structure modulation, which affects honey bee health (Deboutte et al., 2020). Phages are natural structures, safe and well-tolerated by higher organisms, including humans, and can also be safe for bees. Phages exhibiting lytic activity cause destruction and decay only of their bacterial host, without disturbing the composition of the natural gut microflora (Cieplak et al., 2018), and thus they undoubtedly can be applied as therapeutics. Phage ability to amplify at the site of infection is their another advantageous feature, which is why they are called “self-dosing.” Furthermore, it has been suggested that phages may be used in the food industry, preventing the spread of pathogenic bacteria, degradation of food products and also promoting safe environment in animal and plant food production (Sillankorva et al., 2012). Isolation of new therapeutic phages for these purposes is a relatively simple, inexpensive, and rapid process. The use of phages in prophylaxis and treatment of bacterial diseases is a targeted method, with high specificity for the host of antimicrobial activity, less expensive and safer than conventional antibiotic treatments (Fernández et al., 2019). What is more, US Food and Drug Administration (FDA) approved a phage preparation as food additives in 2006, with a status of generally recognized as safe (GRAS) (García et al., 2008; Moye et al., 2018).

Knowledge of the potential use of bacteriophages in the fight against AFB is sparse, and a small percentage of studies devoted to this subject contribute to this situation. Therefore, in this article, we present data concerning phage application against P. larvae infections.

Antibiotics are not fully effective antimicrobials when applied in AFB treatment, they can cause many deleterious effects, and do not destroy P. larvae spores; they treat symptoms, but do not cure the disease, because they prevent the multiplication of only the vegetative forms of bacteria (Genersch, 2010). Application of antibiotics may result in an imbalance in enteric homeostasis, e.g., a disturbance in the influence of honey bee gut microbiota on bee metabolism or immune response, and increase the chances of fungal infection (Raymann et al., 2017). Moreover, when used for a long time, they may cause selection of resistant mutants among different P. larvae strains (Tian et al., 2012; Alippi et al., 2014), leading to antibiotic ineffectiveness. Resistance genes are encoded by mobile genetic elements; bacterial strains acquire them as a result of horizontal gene transfer through phage transduction (Gómez-Gómez et al., 2019). There are data suggesting that antibiotic resistance genes can remain in the environment even for 30,000 years (D’Costa et al., 2011). Interestingly, these genes can be detected in food and transferred to different ecological niches (Godziszewska et al., 2016). Antibiotic residues were detected in different honey bee products, e.g., honey, wax or royal jelly (Hammel et al., 2008; Lopez et al., 2008; Bargańska et al., 2011), thereby reducing honey quality, and potentially affecting the vitality and longevity of bees. A serious threat to humans may be associated with the possibility of antibiotic residue accumulation in commonly consumed bee products (Ortelli et al., 2004; Saridaki-Papakonstadinou et al., 2006).

The World Health Organization (WHO) indicated antibiotic resistance as one of ten biggest threats to global health (World Health Organization (WHO), 2019). Antibiotic application causes the possibility of resistance development in P. larvae, and it has already been detected both in the United States, Canada, and Argentina isolates (Miyagi et al., 2000; Evans, 2003; Alippi et al., 2007). It may be acquired, e.g., via genetic transfer (by mobile genetic elements, e.g., plasmids) even between different bacterial genera from soil, and plasmids encoding antibiotic resistance genes were detected, e.g., in commercial honey (Alippi et al., 2014). Wild strains of P. larvae were proved to carry oxytetracycline resistance genes (Alippi et al., 2007). It has been demonstrated that resistance of wild P. larvae strains may reach even 58% of the samples (Alippi et al., 1999; Miyagi et al., 2000; Spivak, 2000; Murray and Aronstein, 2006; Mitrano et al., 2009). Elzen et al. (2002) reported that macrolide Tylosin Tartrate was more effective in controlling oxytetracycline-resistant P. larvae, with no effect on adult and larval bee mortality.

In the United States, antibiotics are permitted in the elimination of P. larvae. For decades, oxytetracycline was the only approved antibiotic used for this purpose. But since 2005, the FDA has approved four new products to control the disease. For example, the second antibiotic, tylosin tartrate (TYLAN, TYLOVET, TYLOMED-WS), was approved in 2005, whereas the newest antibiotic against AFB – lincomycin hydrochloride (LINCOMIX) – was approved in 2012 (FDA (Food and Drug Administration), 2020).

In the past, antibiotics and sulfonamides were used in EU in the treatment of colonic diseases. Current legislation (Regulation EEC2377/90 and amendments) prohibits the use of antibiotics and does not allow the presence of their residues in honey and hive-derived products, which prevents their application and, in consequence, limits the range of available methods to fight AFB.

Because an increase is observed in the frequency of antibiotic resistance in bacteria, and stringent regulations prohibiting the use of antibiotics in bee disease treatment, phages are suggested as components that may be intended to combat microbial resistance (NIH, 2014) as effective and safe agents in AFB treatment and prevention. Bacteriophages are naturally occurring bacterial viruses that can be found in the hive and honey bee organisms. Phages active against Bacillus larvae were isolated for the first time by Smirnova (1953) from bee larvae suffering from AFB. The source of phages can be: lysogenized bacteria (Gochnauer, 1955; Dingman et al., 1984), water, soil from the hive area (Popova et al., 1976; Valerianov et al., 1976; Ribeiro et al., 2019b), swabs from hive surfaces, beehive materials (Beims et al., 2015) wax, brood (Oliveira et al., 2015), larvae, adult workers and even cosmetics, e.g., containing honey as an ingredient (Merrill et al., 2014; Stamereilers et al., 2016; Yost et al., 2016, 2018; Walker et al., 2018; Tsourkas, 2020). They may be present in the material in which the host bacteria were isolated. Figure 2 presents possible sources of P. larvae phage isolation. The phages showed high specificity for P. larvae and both lytic (Yost et al., 2016) as well as temperate phages were isolated (Dingman et al., 1984). According to some of the recent data, all so far described phages in P. larvae are lytic in vitro (Stamereilers et al., 2018), including those induced from prophages, while other study has indicated that they are all temperate (Tsourkas, 2020). Therefore, this classification may cause incompatibilities, and researchers should be very cautious, for example because phages firstly identified as lytic may prove to be induced from bacteria (i.e., they carry integrase genes in their genomes) after detailed analysis, including BLAST (Stamereilers et al., 2018). Strictly lytic phages are safe when applied in phage therapy, because they do not have the possibility to incorporate into the bacterial genome and transduce bacterial genes when compared to temperate ones (Górski et al., 2020). Temperate phages, especially those capable of transferring antibiotic resistance genes, are not safe and should be excluded from phage therapies. However, there are data suggesting that temperate phages could potentially find use in therapy (Chung et al., 2012; Meader et al., 2013), especially in the fight against AFB, as presented by Ghorbani-Nezami et al. (2015). Of course, for safety reasons, their application should be carefully considered. For example, inability to transduce should be proved at the gene level, as in the study of Ribeiro et al. (2019b) on vB_PlaP_API480, and only then phage application potential can be assessed and confirmed in vivo.

Morphologically, P. larvae phages were mostly identified as an elongated-capsid siphovirus and round-capsid siphovirus (Merrill et al., 2014), but myoviruses were also found (named Abouo, Davies, Emery, Jimmer1, and Jimmer2). The first group of siphoviruses contains phages Diane, Fern, Hayley (Stamereilers et al., 2016), BLA, PBL1, and PPL1c (Merrill et al., 2014). These phages have long, non-contractile tails and elongated capsids. The second siphovirus group contains only PBL3, which have a round capsid (Campana et al., 1991; Merrill et al., 2014). The size of siphovirus phages is approximately 300 nm; phage capsid is approximately 100 nm-long and 50 nm-wide, and tails are approximately 150–200 nm in length (Stamereilers et al., 2016). Myoviruses such as Abouo, Davies, Emery, Jimmer1, and Jimmer2 are similar in size to siphoviruses, with an average capsid height of 67.2 ± 3.2 nm and an average width of 64.1 ± 2.6 nm. The average tail length is about 122.0 ± 27.3 nm (Merrill et al., 2014). Ribeiro et al. (2019b) reported isolation of a Podoviridae phage, vB_PlaP_API480, with activity against P. larvae, characterized using transmission electron microscopy as a phage with an icosahedral capsid and a short 12 × 8 nm non-contractile tail, 58 nm in diameter.

Since the first isolation of phages active against P. larvae, many phages were isolated and described, but only some of them were analyzed for their activity in AFB control. Table 2 presents selected data regarding P. larvae phage application in bees. The available studies fed infected or healthy larvae with phages or sprayed hive elements, e.g., combs. Ribeiro et al. (2019a) investigated in an in vivo study the ability of an active phage to penetrate larvae after per os administration of adult honey bees. T7 phage suspension in 50% (w/v) sucrose was applied and phage biodistribution was assessed in adult bees and larvae; phage penetration through food was confirmed in the larval midgut epithelium, which indicated that phages could be active at the site of P. larvae infection.

Unfortunately, phages, similarly as traditional antimicrobials (e.g., antibiotics), have the ability to destroy only vegetative forms of the causative agent of AFB, they are not able to destruct extremely infective spores.

Phage application in apiaries in vivo should be preceded by their detailed characterization (phage activity, lytic spectrum, life cycle parameters, genome sequences, phage stability under expected conditions at the site of application or infection) and testing their effectiveness in vitro. Ribeiro et al. (2019b) described the API480 phage isolated from a hive soil sample in Spain. The phage showed a broad lytic spectrum and was active against 69% of the tested field P. larvae strains in vitro representing both ERIC I and ERIC II genotypes. The integrase gene and lysogeny module were not identified in its genome. Examination of phage infection parameters revealed that adsorption was achieved several minutes after phage contact with bacterial cells, 85% of phage particles was adsorbed to its host after 35 min. The latent period lasted approximately 30 min., whereas its burst size was 3 pfu per bacterial cell. Furthermore, the phage was proved to be stable in high 50% (w/v) glucose concentration for 24 h and a slight reduction in phage titer (not statistically significant) was observed in homogenized larvae only after 24 h. These features together with the activity (despite the fact that the phage is temperate) observed in vitro suggested that it could be a good candidate for application in hives to treat or protect honey bees in field conditions.

Available studies indicated a possible protective effect for bees infected with these extremely resistant forms of bacteria. The aforementioned effect of three P. larvae phages F, WA and XIII was studied in vivo by Ghorbani-Nezami et al. (2015) on larvae infected with NRRL B-3650 spores. The authors observed that the survival of larvae treated with phages (phage-treated control) as well as healthy larvae (negative control) was comparable and phages did not cause any deleterious effects. Based on these observations, the authors concluded that applying phages as prophylaxis (before infection with spores) provided better results than using them as therapeutic agents (applied after onset of infection symptoms). However, Brady et al. (2017) showed that phage cocktail active against P. larvae may be effective when used both as prophylaxis and therapeutic. Even an overdose cocktail did not exert adverse effects on the mortality of treated bees. The authors compared the efficacy of phage preparations and Tylan Soluble antibiotic, and a 19% decrease in hive health was observed in case of phage treatment applied as a therapeutic, compared to a 38% decrease caused by Tylan. Furthermore, phage application protected hives against P. larvae infection in 100%, whereas 80% of untreated hives were infected. These data indicated the potential of phages, especially in the prevention of AFB infection, and showed that a properly composed phage cocktail can be safer and more effective than antibiotics.

The use of phages or endolysins in the treatment of AFB in honey bee has not only advantages, but also some limitations. Phages are characterized by genomic plasticity. They are able not only to replicate, but also mutate in a specific bacterial host. Additionally, they may induce expression of undesirable virulence factors, toxins and/or antibiotic resistance genes in the host (Łobocka et al., 2014). Most of the isolated phages against P. larvae, such as Davies, Jimmer1, Jimmer2, and PG1 are temperate because they encode integrases or transposases, which excludes their possible application in AFB treatment (Merrill et al., 2014; Stamereilers et al., 2018). Phage therapy against P. larvae, which is a spore-forming bacterium, may fail because of the possibility of spores protecting bacteria against lysis, which may lead to the development of reinfection (Beims et al., 2015).

Bacteria may already be resistant to phages. This situation means that bacterial susceptibility to selected bacteriophages should be tested every time whenever phages are planned to be utilized against these bacteria. An easier and faster solution is to prepare a cocktail that contains phages of different lytic spectra that can be active against wider host ranges, and application of this type of formulation may limit the probability of acquiring resistance to the applied phages (Merabishvili et al., 2018). Furthermore, bacterial collections on which the preparation activity will be tested should be regularly renewed with pathogenic strains from the area where phages are planned to be used to ensure that the phage preparation is active against currently or locally occurring bacterial strains (Merabishvili et al., 2018). Another possibility is to try to isolate new phages from materials from which pathogenic bacterial strains (that caused AFB) are isolated or various environmental samples that are collected.

There are many factors that may interfere with phage activity (when phages are intended as therapeutics), such as physico-chemical conditions, host physiological conditions, preparation composition or phage inability to penetrate and achieve high concentration at the site of infection (Jończyk-Matysiak et al., 2019). Phage efficiency may depend on its properties, structure and biology as well as therapeutic expectations, dose, manner of application, as well as modifications that can improve their activity and availability. The individual stability of the phage at different pH conditions is also an important issue, and this feature should be checked for each phage, as they tend to have different sensitivities to various physico-chemical conditions (Jończyk-Matysiak et al., 2019). These features should be taken into account and all phages contained in the preparation should be fully characterized. In addition to phages present in the preparation, it is also important to select proper additives that would protect phage activity and act as stabilizers. Moreover, the entire composition of the preparation intended for use in hives or to feed bees should be well tolerated by bees (taste and safety) and contain as few ingredients as possible.

Penetration of phages to the honey bee gut – the site of P. larvae infection – should be tested to ensure that the phages penetrate and are active at the site where AFB etiological agent is present. Therefore, spraying hive elements seems to be a less effective route of phage administration, resulting in a lower phage concentration on the hive surface. Despite the confirmed phage penetration at the site of infection, limitations of their action may be associated with low phage concentrations (as observed by Ribeiro et al., 2019a), not sufficient to reduce the count of P. larvae to prevent and cure AFB. Therefore, a high dose of phage particles should be provided and a method of phage protection against harmful hive-derived conditions is required (e.g., temperature, humidity, pH of larval food, persistence on the surface of hive elements). That is crucial to retain phage activity, as the conditions listed above vary in hive throughout the year. Bees have mechanisms to control the nest climate and specifically the brood area, but they depend on the prevailing weather conditions outside the nest and bee colony metabolism, which changes during the year (Stabentheiner et al., 2003, 2010; Tautz, 2008; Shaw et al., 2011; Cecchi et al., 2020). Moreover, phage preparations applied in the winter can remain deposited in wax patches even for several months and can cause significant reduction in phage activity. Under varied temperature conditions, phages may lose their activity during persistence in wax patches (Weinbauer, 2004); therefore, in order to protect phages applied in hives to control AFB, methods harmless to bees prolonging phage activity, such as encapsulation providing protection against harmful external factors, chemical or genetic modifications enabling extended phage activity, and addition of stabilizers protecting phages against their titer reduction should be considered.

It was demonstrated that phages could be inactivated during storage in larval food for a long time (Gochnauer, 1970). It was suggested especially in the case of larvae fed with phages and royal jelly that phage inactivation could be caused by the latter (low pH, composition), which could be one of the possible causes of failure of phage therapeutic effect in bees suffering from AFB (Yost et al., 2016; Ribeiro et al., 2019a). Hence, phages should be taken by larvae as fast as possible to ensure a neutral pH in gut conditions that guarantees phage stability. In addition to royal jelly, honey was proved to completely inactivate phages in vitro (Oliveira et al., 2017, 2018), which indicated that hive-derived products could influence phage activity in the hive environment.

Although in vitro studies on phages or their lysins are promising, phage activity depends on hive environment, larval gut conditions, phage individual features and its stability may significantly limit the effects of phage application in vivo. To protect phages and lysins against inactivation in hive and maintain phage stability during and after phage application, different methods may be used, e.g., encapsulation, addition of stabilizers, chemical or genetic modifications (mentioned in section “Endolysins and Their Potential Against P. larvae”) to achieve prolonged activity or render phage or its protein extremely insensitive to environmental conditions.

Lysins appear to be effective against specific bacteria. However, studies using P. larvae phage lysins are scarce, which makes it difficult to infer from them meaningful conclusions. Since 2015, two lysins have been tested that (similarly to phages) were neither active against dormant nor germinating spores (Le Blanc et al., 2015; Oliveira et al., 2015). More studies should be conducted on a larger number of phage lytic enzymes in both laboratory and hive conditions, especially to allow comparison of the effects that form the full range bactericidal activity of P. larvae-encoded lysins. Moreover, studies should be undertaken to develop methods allowing to make lysins more effective (especially against spores), active at the site of infection – in the larva’s gut.

Currently, the trend of using natural products continues both in the treatment of bacterial infections and as part of a diet. Research by Naanwaab et al. (2014) showed that consumers declared to pay extra for bacteriophage-treated fresh product if it would improve their food safety. This indicated that consumers were not afraid of phage application in food, which can suggest that phage residues in bee products may also be acceptable.

Therefore, further research on both bacteriophages and lysins in the fight against AFB is required. Research in this area is still very limited and is urgently needed to save billions of honey bees all over the world.

There is an urgent need for a safe, natural and effective product for the prevention and treatment of AFB in honey bees, whose application would not cause any adverse effects. There have been reports of isolation of bacteriophages active against P. larvae, and attempts of their use in AFB prevention and treatment. Published data indicated that the isolated specific phages showed the ability to lyse only the vegetative forms of P. larvae strains, and in vivo effects suggested that phages could be particularly useful in AFB prevention rather than treatment. Studies on phage-encoded endolysin are also promising. The use of phages or their enzymes in AFB therapy may reduce the need to eliminate hives by burning. Further research on phage application using different phage titers, different phage formulation compositions, forms of preparation (lysate or purified preparation), application (feeding, spraying), addition of different carriers both in vitro and in vivo during different stages of bee development is required. In addition, their application may reduce the risk of negative influence on the health of bees and consumers of bee products. Therefore, the possibility of using phages in treating AFB can bring great economic and environmental benefits as well as advantages for human health.

EJ-M conceptualized the study and drafted the main part of the manuscript. EP, BO, KH-S, KŚ-J, NŁ, DK, JN, PM, NB, FO, BW-D, AR, and AG contributed to the parts of the manuscript. EP and NŁ prepared figures. EJ-M, EP, and KH-S prepared tables. AG provided support and conceptual advice at all stages of manuscript preparation. All authors revised the manuscript.

BW-D and AG are co-inventors of patents owned by the L. Hirszfeld Institute and covered phage preparations. KH-S and KŚ-J were employed by the company Pure Biologics. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.