Listeria phages

Abstract

Listeria is an important foodborne pathogen and the causative agent of Listeriosis, a potentially fatal infection. Several hundred Listeria bacteriophages have been described over the past decades, but only few have actually been characterized in some detail, and genome sequences are available for less than twenty of them. We here present an overview of what is currently known about Listeria phage genomics, their role in host evolution and pathogenicity, and their various applications in biotechnology and diagnostics.

Listeria are Gram-positive, rod-shaped members of the Firmicutes, currently divided into nine species, namely L. innocua, L. ivanovii, L. monocytogenes, L. grayi, L. seeligeri, L. welshimieri, L. marthii, L. rocourtiae, and L. fleischmannii.^1-5 Only L. monocytogenes are human pathogens, causing the rare, but life-threatening disease Listeriosis, which mainly affects young, old, pregnant and immunocompromized individuals, and is associated with mortality rates of up to 30%. The disease can manifest as septicemia, meningitis or meningoencephalitis; or result in stillbirth or abortion in pregnant women.⁵Listeria is ubiquitously found in the environment, and transmission usually occurs via contaminated food and water. The organism can grow at low temperatures and in high-salt environments, and is also known to readily form biofilms.^6,7 These properties render it a high-risk organism for food producers.

Listeria Bacteriophages

Bacteriophages are the natural enemies of bacteria. Recently, they have again moved into the focus of research interest, with respect to biocontrol of pathogenic bacteria as well as offerering tools for novel and effective separation technologies and diagnostics. In fact, bacteriophages present ideally suited means to control and detect Listeria cells in foods. Listeria phages can be isolated with relative ease from various environmental sources by the soft agar overlay method.⁸ First reports on bacteriophages specific for Listeria monocytogenes date back to the 1940s and 1960s.^9-12 To date, more than 500 Listeria phages have been isolated and characterized to a certain extent, most of them in the course of phage typing studies. All Listeria-specific bacteriophages found to date are members of the Caudovirales, featuring the long, non-contractile tails of the Siphoviridae family, or the complex contractile tail machines of the Myoviridae family. Interestingly, no podoviruses have ever been isolated for Listeria spp. The reason for this is unknown, but might be associated with the structure of the Listeria cell; yet, morphological diversity in the Listeria phages seems limited. The Siphoviridae are currently grouped into six species, depending on tail length. The presence of intact or cryptic prophages has been confirmed in many Listeria strains, e.g., phage A118 in L. monocytogenes WSLC 1118 or PSA in L. monocytogenes ScottA.^2,13-15 Bacteriophages have been found in all major Listeria species and serovars, except the unusual L. grayii, the newly proposed species L. rocourtii and L. marthii, as well as in serovar 3 strains. In general, serovar 3 strains are highly refractory to phage infection, whereas serovar 4 strains are very sensitive. Interestingly, strains of both serovars 4b and 4c also seem to lack any intact prophage genome.¹⁶ Prophage absence in serovar 4 seems to be attributable to differences in teichoic acid composition.¹⁶ Although being very sensitive to phage infection, also no prophages have been found in strains of L. ivanovii subsp ivanovii.¹⁷

Temperate Listeria phages feature a generally rather limited host range, partially due to homoimmunity, since the majority of Listeria strains carries prophages or prophage remnants. This is in contrast to the broad host range Listeria Myoviridae, which can be grouped in two clusters, the A511-like phages of the newly formed subfamily Spounavirinae, and the B054 and 01761-like phages, both of which feature a slightly larger genome (48.5 kb) then the Siphoviridae members.

Table 1 presents an overview of the best characterized Listeria bacteriophages. Many more have been isolated but are poorly characterized (if at all), including 25 phages from the international typing set, seven “experimental” phages,¹⁸ fourty-two previously not described phages used by Hodgson during a transduction study¹⁹ and 114 Listeria phages from silage of two dairy farms in New York State, USA.²⁰ Vongkamjan et al. report a host-range for all phages. Serovar groups 4 and 1/2 were found to be particularly susceptible to phage infection, while the other Listeria serovars were lysed to a varying degree. No phages have been reported for Serovar 3 strains, confirming earlier observations. Surprisingly, a very large proportion of the isolated phages feature genome sizes between 57–68 kb, in range of the 67 kb found for the recently reported novel phage species P70, featuring an elongated head morphology.^20,21 This would suggest that the novel phage species is abundant in environmental samples, possibly possessing a selection advantage over their smaller genome siphovirus relatives. A similar, although smaller-scale study reported the isolation of 12 Listeria phages from a Turkey processing plant in the United States.²² Arachchi et al. reported isolation of three phages active against L. monocytogenes from seafood²³; Zhang et al. described phage FWLLm3, of which only the endolysin sequence is known²⁴; and Anany et al. introduced three new Listeria phages (LmoM-AG8, 13, and 20).²⁵ The LmoM-AG20 genome is very similar to A511 and P100 (H. Anany, personal communication). Many more preliminary descriptions of Listeria phages and their isolation exist, highlighting their ubiquitous distribution and underlining the tremendous interest in these bacterial viruses.

Table 1. Synopsis of characterized Listeria bacteriophages.

Phage name	Family (EM-confirmed)	Dimensions (head × tail)	Genome size (bp)	Lifestyle (host name)	Host serovar	Remarks	Refs.
01761	Myoviridae	66 × 270	48306	Temperate (PS1803)	1/2		84, unpublished data
11355C	Siphoviridae	63 × 320			1/2		84
11711A	Siphoviridae	63 × 320			1/2		84
02971A	Siphoviridae	64 × 307			1/2		84
02971C	Siphoviridae	64 × 306			1/2		84
90666	Siphoviridae	60 × 311			4		19, 84
907515	Siphoviridae	60 × 295			1/2		84
10072	Siphoviridae	61 × 292			1/2		84
12981	Siphoviridae	63 × 302			1/2		84
13441	Siphoviridae	61 × 315			1/2		84
00241	Siphoviridae	64 × 300			1/2		84
00611	Siphoviridae	63 × 301			1/2		84
90861	Siphoviridae	61 × 173			4		19, 84
910716	Siphoviridae	60 × 179			4		19, 84
93253	Siphoviridae	61 × 171			4		19, 84
A005		62 × 280			1/2		84
A006	Siphoviridae	62 × 280	38124	Temperate (WSLC 1006)	1/2	Transducing	19, 30, 31, 33
A020	Siphoviridae	63 × 248		Temperate (WSLC 1020)	4, 5		30, 50
A118	Siphoviridae	61 × 298	40834	Temperate (WSLC 1118)	1/2	Transducing	13, 19, 30
A500	Siphoviridae	62 × 274	38867	Temperate	4	ATCC 23074-Bl, transducing	19, 30, 31, 33, 50
A502	Siphoviridae	62 × 302	~39000	Temperate	1/2	Isolation from sewage, transducing	19, 30–32, 50
A511	Myoviridae	87 × 199	137619a	Virulent	1/2, 4, 5, 6	Isolation from sewage	19, 27, 30, 32, 50
A640	Siphoviridae	62 × 305		Temperate	4		19, 84
B012	Siphoviridae	61 × 286	41464	Temperate (WSLC 2012)	5, 6		30–32, 50
B021	Siphoviridae	61 × 302		Temperate	4		19, 84
B024	Siphoviridae	59 × 239	~37000	Temperate (WSLC 2024)	5, 6		30–32, 50
B025	Siphoviridae	63 × 252	42653	Temperate (WSLC 2025)	5, 6		30–33, 50
B035	Siphoviridae	63 × 294	38881	Temperate (WSLC 2035)	5, 6		30–32, 50
B051 (4211)c	Siphoviridae	62 × 245		Temperate (WSLC 2051)	5, 6		30, 32
B053	Siphoviridae	59 × 244	43471	Temperate (WSLC 2053)	5, 6		30–32, unpublished data
B054 (4286)c	Myoviridae	64 × 244	48172	Temperate (WSLC 2054)	5, 6		30–33
B055 (4295)c	Siphoviridae	62 × 242		Temperate (WSLC 2055)	5, 6		30, 32
B056 (5337)c	Siphoviridae	59 × 285	~35000	Temperate (WSLC 2056)	5, 6		30–32
B101	Siphoviridae	61 × 280	40862	Temperate (WSLC 2101)	5, 6		30–32, unpublished data
B110	Siphoviridae	57 × 288	39390	Temperate (WSLC 2110)	4, 6		30–32, unpublished data
B545	Siphoviridae	62 × 258		Temperate (WSLC 2545)	5, 6		30, 32
B620		61 × 299					84
B640	Siphoviridae	62 × 305		Temperate (WSLC 2640)	4	Transducing	19, 84
B653	Siphoviridae	61 × 260	37943	Temperate (WSLC 2653)	1/2, 4, 5, 6		30, 32, unpublished data
C707	Siphoviridae	60 × 243			5	Isolation from sewage	30, 32, 50
D441	Siphoviridae	63 × 247	~37000	Temperate (WSLC 4441)	4, 5		30–32
P35	Siphoviridae	58 × 110	35822	virulent	1/2	transduction experimentally proven	19, 33
P40	Siphoviridae	56 × 108	35638	Virulent	1/2, 4, 5, 6		33
P70	Siphoviridae	(128 × 57) × 141	67170	Virulent	1/2, 4, 5, 6		21
P100	Myoviridae	90 × 198	131384b		1/2, 4, 5, 6		26, 27
PSA	Siphoviridae	61 × 180	37618	Temperate	4		14, 19, 84

^a Including terminal redundancy of 3125 bp. ^bUnit genome size, probably features a 6 kb terminal redundancy. 
^cNumbers in parentheses are designations by Rocourt et al.⁸⁵

Listeria Phage Genomics and Implications for Phage Evolution

Almost all Listeria phage genomes described today feature sizes between 30 and 65 kb, with only a handful of larger genomes (125–140 kb), featured by the broad-host-range myoviruses.^26-28 All Listeria bacteriophage genomes are composed of dsDNA and are organized in a modular fashion, usually comprising a module which encodes structural proteins, a module of early genes, encoding functions for DNA recombination, replication and repair, a lysis cassette featuring a holin and an endolysin gene and, in case of temperate phages, a lysogeny control region. Open reading frames of the latter are usually transcribed in opposite direction compared with the other genes. As it is the case with many other sequenced bacteriophage genomes, a large fraction of open reading frames encode products with weak or no homology to any other proteins. Approximately 50% of predicted phage proteins have no known function, and have been postulated to assume roles in host takeover, nucleotide metabolism, and transcription regulation.

Two large myoviruses infecting Listeria cells have been characterized in molecular detail, A511 and P100. Their genomes are very similar, with P100 featuring a 3 kb smaller unit genome, but a larger terminal redundancy of approximately 6 kb, compared with the 3.1 kb termini of A511.²⁷ P100 and A511 are morphologically indistinguishable, which is also reflected by a highly similar genome organization and strong sequence homologies among structural proteins. Both phages feature a baseplate structure which during tail contraction undergoes a structural rearrangement into a double-ringed organelle, a hallmark of the Twort-like phages of the Spounavirinae subfamily.²⁸ Both short and long tail fibers are present and involved in host recognition and adsorption. The A511 baseplate is among the most complex contractile injection systems known in nature,²⁹ and current research in our lab aims to elucidate the fine details of this superstructure. The major components of the adsorption apparatus feature strong sequence and secondary structure homologies to other phages of the Spounavirinae (Habann et al., submitted). Interestingly, two temperate myoviruses of Listeria have been described, namely B054^30-33 and 01761.³⁴ Both feature genes responsible for host genome integration and maintenance of a temperate lifestyle (ref. ³³ and unpublished data).

Almost all known Listeria phages belong to the Siphoviridae, while no members of the short-tailed Podoviridae have been found or described. Most Listeria phages are classical lambda-type siphoviruses, such as A118 or PSA, which feature a long, flexible tail and an isometric head. Typical genome sizes range from 35 to 40 kb. All members of this type of phages are temperate, i.e., they encode a lysogeny control region, including an integrase gene in their genome and host strains can be lysogenized, or the phage has been be induced from lysogenic carrier strains.^13,14,33,35 Still, morphology and genomes can be strikingly different even among the siphoviruses. The morphologically similar, but distinct phages P35 and P40 feature a lytic lifestyle and have a relatively broad host range among specific serovar groups. We also found that they do not possess the highly complex adsorption apparatus of myoviruses, their tail is not flexible but also not contractile, and their genomes lack genes associated with in lysogeny control.^19,33 They have been proposed to possibly represent archetype phages in evolutionary transition from a lytic (i.e., virulent) to a temperate lifestyle.

The newest addition to the Siphoviridae group are phages featuring the B2 capsid morphotype, such as the recently described phage P70.²¹ The elongated head holds a larger genome of about 65 kb, and the phage features no relationship to the other known Listeria phages. Vongkamjan et al. also reported isolation of Listeria phages with similar sized genomes, suggesting that they could possibly fall into this new species.²⁰ As described for P35 and P40 above, P70 also features a virulent lifestyle and an unusually broad host range. This is in agreement with the lack of genetic determinants indicative of a lysogeny control module.²¹

Lysogeny is widespread in Listeria¹⁷ and the presence of prophage does not correlate with epidemic clones. Some strains, such as L. innocua CLIP11262 can harbor up to six prophage(-like) elements, whereas L. monocytogenes EGD-e harbors only two prophage elements, and other strains (such as serovar 3 strains, WSLC1001 and WSLC1042) seem to be free of intact prophage sequences.^2,36-38 Temperate phages can be induced from lysogens using UV light or mitomycin C,³⁵ and lysogens are resistant to infection by the same or related phage.³⁶ Many of the temperate Listeria phages are capable of generalized transduction,¹⁹ where random fragments of host DNA are packaged into the phage head during assembly. These particles are infective and can inject the bacterial DNA into susceptible cells, but can obviously not complete the infection cycle. However, genes for virulence factors, toxins and antibiotic resistance can be transferred. The ability to perform generalized transduction is dependent on the physical genome structure of the phage. Phages with circularly permuted genomes, such as A118,¹³ can perform transduction because the terminase enzyme exhibits no or only low sequence specificity and packs the proheads using a “headful” mechanism. In contrast, phages with fixed invariable genome structures, such as PSA or A511,^14,27 are unable to transduce DNA because the terminase recognizes a specific sequence on the DNA, and this recognition event is crucial for DNA packaging. The absence of transduction is an important prerequisite for biotechnological application of phage preparations to kill viable bacteria (see below). The majority of known Listeria phages features a circularly permuted genome structure (Table 1), a simple structure which allows loss-free DNA replication by circularization upon cell entry.³⁹

Temperate Listeria phages encode integrase enzymes which mediate integration of the phage genome into the host chromosome, and excision of the phage genome under inducing conditions.⁴⁰ Depending on the specific catalytic amino acid residue present, phage integrases are classified as serine or tyrosine recombinases.⁴¹ Phage A118 features a serine integrase, which recognizes a sequence with only three basepairs homology between phage and bacterial attachment site.¹³ In contrast, the PSA integrase is of the tyrosine type, and recognizes a 15 bp site located at the 3′ end of a tRNA_Arg gene.¹⁴ Phages A006, B025, and B054 also encode recombinase enzymes, recognizing 16–17 nt core sequences, whereas the attachment and recombination sequence required for the recombinase from phage A500 is 45 bp long. All four of these enzymes are presumably of the tyrosine type. Both B025 and PSA integrate in the same locus, a tRNA_Arg, excluding simultaneous presence of both prophages in the same host cell. A006 and A500 target other tRNA genes.³³ It was recently speculated that tRNA genes represent “anchoring elements” for prophage uptake in Listeria.³⁸ In contrast, the recombinase recognition site for B054 is different and quite unique, since it integrates into a transcription elongation factor.³³

Albeit the availability of approximately 50 Listeria phage genomes and a growing number of host genomes, no clear indication for an influence of bacteriophage on Listeria pathogenicity or virulence has been found. Recently, however, an A118-like¹³ prophage of Listeria monocytogenes integrated into the comK gene has been shown to play a role in the regulation of phagolysosome escape of the organism during infection.⁴² Until then, no function could be assigned to the Listeria comK, a homolog of the natural competence regulator in Bacillus subtilis. Phage integration disrupts the comK gene, whereas phage excision restores the reading frame.^13,43 An earlier study suggested a role of a comK-integrated prophage in the persistence of L. monocytogenes in food processing plants, attributing higher cell densities after 48 h incubation on meat to comK-prophage- containing lysogens.⁴⁴ However, this observation has never been supported by studying possible underlying mechanisms, such as the increased formation of biofilms due to a higher fraction of lysed cells and better availability of nutrients, and other effects.

Recently, genome and transcriptome studies of L. monocytogenes found prophage gene expression to be a hallmark of intracellular gene expression. The expression of phage-derived genes is upregulated when Listeria proliferates intracellular. In contract, deletions in a single shared prophage (the “monocin locus”—see below) of two lineage II strains led to severe virulence attenuation.³⁸ Unfortunately, the underlying mechanism(s) leading to this attenuation remain yet unclear.

Last, but not least, a putative CRISPR system⁴⁵ has also been identified in Listeria and its phages. It is composed of two adjacent loci with a considerable difference in the number of repeats; locus II is only present in serovar 4b strains, while locus I is conserved in serovar 1/2a and 4b strains, and in L. innocua. Not all strains feature identifiable cas genes,³⁸ and only locus II seems to be functional.¹⁶ A third repeat locus was recently identified in serovar 1/2a, 1/2b, 3b, and 7 strains and contains cas gene homologs.¹⁶ However, the presence or absence of such typical CRISPR elements does not correlate with the presence of a certain type or number of prophages in Listeria, and it can therefore only be speculated if CRISPR/Cas systems provide defense against bacteriophage infection in Listeria.¹⁶

Applications of Listeria Phages

Listeria poses a severe health threat to the consumer. Its unique ability to survive high salt concentrations and proliferate at refrigeration temperatures makes it especially difficult for food producers to eradicate Listeria from the production chain. Applicable law in most countries requires that Listeria monocytogenes is absent from 25 g of a food sample, which obviously is a challenge for any detection system. Given its sporadic occurrence on various types of food, its ubiquitous distribution in nature, the high mortality rate in infected individuals and its unknown infection dose, a strong need exists for highly sensitive and rapid diagnostics, as well as efficient and food-grade control strategies for Listeria monocytogenes in ready-to-eat foods. Initially, Listeria bacteriophages have been used for phage-typing Listeria isolates, either from foodborne outbreaks and epidemics, or for source tracking in contaminated production plants. The procedures to be used were devised more than 25 years ago. Several more or less defined typing sets were developed,^46,47 and subsequently supplemented and standardized.^18,30,48-50

Attempts to harness the biological specificity of phages for Listeria host cell detection involved the construction of a luciferase reporter phage A511::luxAB,⁵¹ and its testing in a food environment.⁵² For this purpose, the Vibrio harveyi bacterial luciferase genes luxAB were inserted downstream of the A511 major capsid gene under the control of its own strong promoter. When a Listeria cell is infected, luxAB is transcribed, the luciferase enzyme produced in the infected cells, and in presence of its oxidizable aldehyde substrate (Nonanal) generates a strong and real-time detectable light emission. The assay is sensitive enough to detect one cell per gram of food after a shortened enrichment procedure, which provides a three day time advantage over conventional enrichment and plating protocols.⁵²

The Listeria reporter phage concept has recently been further developed by the construction of A511::celB. The celB gene from Pyrococcus furiosus encodes a thermostable β-galactosidase, which can be used with various chromogenic, fluorescent or chemiluminescent substrates, also creating a detectable response upon phage infection.⁵³ The authors demonstrated the feasibility of the assay in detecting as low as 10 cfu/g/ml of food in chocolate milk and salmon⁵³ using an inexpensive substrate and a simple endpoint assay.

Quite naturally, bacteriophages also represent perfect tools for biocontrol of bacteria, in this case Listeria. In this case biocontrol is most effective by removal of the organisms from contaminated food material, since the pathogen prefers an intracellular lifestyle and phage-based product cannot reach the bacterium once inside the human host.

Biocontrol approaches for Listeria (and any other pathogen) require a number of specific characteristics from the phage(s) to be used: It must be strictly virulent, feature a broad host-range, unable to perform generalize transduction, and does not affect virulence or pathogenicity or result in lysogenic conversion of its host. Ideally, the phage can be propagated on a non-pathogenic host, to avoid handling large-scale pathogen cultures and to prevent any interference of phage preparation impurities with downstream pathogen detection systems.^54,55 Phage P100 was characterized by Carlton in 2005²⁶ and subsequently developed into a product, which has received GRAS-status by FDA/USDA for use in all food materials (2007). The phage has also been used in a number of studies showing its efficacy in removing Listeria contaminations from fish, removing Listeria biofilms from stainless steel surfaces,^56-58 and prevention of Listeria on cooked ham.⁵⁹

A511, a close relative of P100,²⁷ was used in several proof-of-concept studies and was shown to very effectively reduce Listeria contaminations on various ready-to-eat foods by more than 5 log units.⁶⁰ Bigot et al. successfully used an A511-like phage for biocontrol of Listeria in ready-to-eat poultry products, demonstrating up to 7 log reductions in viable counts.⁶¹

Other anti-Listeria phage products exists, such as a cocktail of six phages, which has also been FDA- and USDA-approved for application on food and surfaces in food production facilities. The phage cocktail was used to reduce Listeria occurrence on fresh-cut produce,⁶² and for spray application of phages on melons.⁶³

Listeria phages have also been able to inhibit Listeria growth on artificially contaminated samples,²⁵ even when immobilized on cellulose membranes and used for packaging of ready-to-eat and raw meat products.

Zink and Loessner have described the presence of lytic particles in Listeria cultures, termed monocins.^64,65 These substances are produced from up to 70% of all Listeria cultures and they inhibit other Listeria. Monocins resemble phage tail structures and result from the presence of incomplete, cryptic prophages. The tail associated lytic proteins (used during infection for cell wall penetration) are toxic to certain Listeria species and act as biocins. The “monocin locus”, a cryptic prophage region including the lma genes is conserved in all L. monocytogenes lineages and also present in L. innocua.^38,66 Many monocine-like substances are found in other genera of bacteria, e.g., the “pyocins” of Pseudomonads.⁶⁷ Currently, renewed interest in monocin/pyocin-like substances for biocontrol of pathogenic bacteria is evident.^68-70

Another biocontrol and detection strategy is using recombinant phage-encoded peptidoglycan hydrolases (endolysins) instead of infective virus. The potential of bacteriophage-derived endolysins in controlling foodborne pathogens has been subject of several reviews, e.g., ref. 71. However, systematic studies are lacking. Listeria phage endolysins bear potential as disinfection agent, similar to what has been done in control of Streptococcus equi⁷² and other Gram-positives.^73,74 The regulatory approval for protein based disinfectants may be less challenging than approval of a virus-based food additive.

Over the past decade, molecular details of Listeria phage endolysins have been elucidated. The crystal structure of two of them is known,^75,76 and their generally modular composition and binding ligands on target cells are intensively studied.^77-81 Cell wall binding domains of the endolysins have been used to decorate Listeria with fluorescent labels, in order to discriminate them in mixed bacterial culture.⁷⁸ These CBD binding domains have also been used for immobilization of Listeria cells, following coating on paramagnetic beads and specific capture of Listeria cells from milk and other matrices, followed by plating or real-time PCR quantification.^82,83

In conclusion, bacteriophages do not only play an important role in Listeria genome plasticity and evolution, but offer a large and versatile toolbox for development of novel detection and biocontrol methods for Listeria, based on infective viruses, genetically modified phage, and native or recombinant phage-encoded enzymes.

Citation: Klumpp J, Loessner MJ. Listeria phages: Genomes, evolution, and application. Bacteriophage 2013; 3:e26861; 10.4161/bact.26861

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.