Over 99% of phages detected using microscopy have not been cultured. This article explores factors that influence plaque formation and if addressed may help in phage isolation.
Current data indicate that some 1031 bacteriophages exist globally, including about 108 genotypes. Some phages form very tiny or micro plaques. These can sometimes be so small that it is almost impossible to see them. Frequently 'new' phages can be observed using e.g. electron microscopy under conditions where there is strong evidence of a potential host yet it can be very time consuming or in some instances not possible to get the phage to form plaques. Less than 1% of the phages observed using microscopy have ever been grown in culture, this is sometimes called "the great plaque count anomaly".
The conditions required to get a newly isolated phage to form plaques have been reviewed. The importance of testing both logarithmic and stationary phage cells, a range of incubation temperatures, replacing agar with agarose, using low strength agarose top agar overlays in initial experiments, media supplemented with Ca++ and Mg++ that do not contain cation chelators, modifications to the double agar assay method including a) changes to the initial assay step so that phage adsorption takes place at ambient temperature and b) the use of antibiotics and other activators of the host cell’s SOS system are discussed. The use of metagenomics to study phage diversity and dynamics is also discussed.
The review is based on a papers given by the author (Mullan, 1979; Mullan, 2012) that have been 'broadened' to include non-lactic phages and more recent research. Further updates were added in 2013, 2014 and 2016. Note this article does not discuss how to undertake a plaque assay. For information on how to plate phage preparations refer to the article on the double agar assay method.
Difficulties in getting phages to form plaques
Sample preparation prior to analysis
Replacing the gelling agent, agar, with agarose in the soft agar overlay
Concentration of agarose in the soft agar overlay
Need for the individual component strains as hosts in plaque assay
Use of indicator strains or heterologous hosts in phage assay
Growth phase of host
Influence of growth medium
Broth or milk-grown cells in phage assays
Incorporation of glycine in growth media
Incorporation of sodium azide in growth media
Use of antibiotics that activate the host cell SOS system to increase plaque size
Oxygen status during incubation
Engineering phage sensitive indicator strains?
Using phage metagenomics to identify environmental phages
Phage enumeration using the double agar assay is an essential control test in many quality assurance systems involving bacterial fermentation processes. However, increasingly phages are being used in the treatment of human and animal diseases, to prevent spoilage of fruits and vegetables, to eliminate biofilms in food processing plants, in studies of nano-machines and nanotechnology and in extensive genetic and molecular biology research. The genetic material present in phages is now understood to be a major, relatively untapped, reservoir of the planet's gene diversity. Current data indicate that some 1031 bacteriophages exist globally, including about 108 genotypes e.g. Williamson et al. (2005).
Some phages form very tiny or micro plaques. These can sometimes be so small that it is almost impossible to see them. Frequently 'new' phages can be observed using e.g. electron microscopy under conditions where there is strong evidence of a potential host yet it can be very time consuming or in some instances not possible to get the phage to form plaques. Less than 1% of the phages observed using microscopy have ever been grown in culture, this is sometimes called "the great plaque count anomaly" e.g. Serwer et al. (2007). Without plaques, and ultimately, pure phage, it can be challenging to quantify phage concentrations in the environment and to undertake phage research.
While an overview of the double agar method and the process of plaque formation has been discussed our understanding of plaque formation has been significantly enhanced by the work of Abedon and co-workers who have added to existing research on the mathematical modelling of plaque production. Abedon and Yin (2009) have revealed the importance of phage diffusivity (the rate of virus particle diffusion in the absence of the host), the extent of phage-bacterium attachment, phage latent period, burst size, and host density on plaque production and overall size. This work has now reached the point where it can be applied to improve plaque visibility and to provide a theoretical understanding of why some techniques e.g. the use of antibiotics that activate the host cell SOS system markedly increase plaque size.
This article provides information on factors influencing plaque formation by bacteriophages and will be periodically updated to reflect recent research findings.
The need to prepare samples prior to phage analysis has been discussed. In particular consideration should be given to the presence of host growth inhibitors in samples and their neutralisation if required. Additionally, consideration should also be given to the selection of membrane filters including their apparent size exclusion limits to minimise phage titre reduction effects; some phages may absorb to the filter matrix.
Some phages, particularly very large viruses, may aggregate significantly reducing their titre after either low speed centrifugation or membrane filtration (Serwer et. al., 2007). This is relatively easy to check if researchers have direct access to electron microscopy facilities but is difficult to do otherwise.
Agar is a mixture of polysaccharides some of which may contain both host-growth and virus inhibitors. It is widely used as the gelling agent in plaque assays. However, there is increasing evidence of problems using agar in virus assays. Agarose is purified agar in which most of the agaropectin, the component containing most of the sulphate and carboxyl groups which are considered responsible for inhibition of viruses, has been removed.
Many phages, particularly lactococcal phages, plaque well regardless of whether agar or agarose is used as the gelling agent in the overlay layer. However, plaque size may be improved if the agar is replaced with agarose for some phages. It would seem prudent to use agarose as the gelling agent in studies with ‘new’ phages.
Phage diffusivity has been shown to be related to plaque size (Abedon and Yin, 2009). In general, plaque size increases as the velocity of phage diffusion increases. The diffusion rate is dependent on certain phage properties e.g. phage dimensions and whether the phage aggregates. It is also dependent on the concentration of agar in the overlay layer.
So reducing the agar concentration in the overlay would be expected to increase plaque size for most phages. The critical importance of agarose concentration on plaque formation has been demonstrated for a very large Mycophage, 0305φ8-36, for Bacillus thuringiensis. Serwer et al. (2007) found that the Bacillus thuringiensis phage on initial isolation made small (<1 mm) plaques in a 0.4% agarose overlay but the plaques became progressively larger as the agarose gel concentration decreased to 0.2%. Post isolation, the 0305φ8-36 phage would only form plaques providing the agarose concentration was 0.25% or less (Figure 1). Based on these findings it would seem prudent to consider using 0.2% agarose in screening studies for hosts for new phages. In marked contrast the coli-phage T4 produced plaques at all agarose concentrations tested. However, plaque size did decrease as the agarose concentration increased but levelled off approximately at around 0.2% to 0.6% agarose (Figure 1).
Many phages have an obligate requirement for specific divalent cations, or other cofactors, to propagate or form plaques. While others may grow without the addition of supplementary cations, plaque size can often be increased by their addition. The mycobacteriaphage, L5, will form plaques on un-supplemented media if Mycobacterium smegmatis is used as host. However, these plaques are small and tend not be distributed evenly across plates. Addition of 1 mM CaCl 2 to media results in larger more homogeneously sized plaques. L5 requires calcium ions to plaque on slow growing mycobacteria e.g. Mycobacterium tuberculosis BCG (Fullner and Hatfull, 1997) demonstrating not just the critical importance of cations but that the cation requirement is phage-host dependent.
Phages that require cations e.g. Ca++ may need the cation for nucleic acid injection or introduction (e.g. Paranchych, 1966) or for efficent adsorption to cell wall binding sites (e.g. Fullner and Hatfull, 1997). Additionally some phages may show enhanced stability in the presence of certain cations.
Consequently phage assay media are often supplemented to give 1-10 mM solutions of calcium and magnesium salts. In some phage/host systems while cations are required there is no requirement for a particular cation e.g. the Leviphage R17, a single-stranded RNA phage, requires Mg++, Ca++, Sr++, or Ba++ to propagate but Mn++, Zn++, Ni++, or Co++ do not support growth.
Need for the individual component strains as hosts in plaque assay
This section applies to situations in which workers do not have access to the component strains of a starter but have the starter culture as supplied by the culture supplier. There is little value in attempting to quantify phage concentrations in cheese whey and other environmental samples using host inocula containing more than one strain - the majority of cultures available from culture suppliers. While it is difficult to say the precise level of host that is required for plaque formation in a mixture, plaque formation generally requires the presence of about 20% host and in some instances 40% host in the inoculum used (Zameni et al., 1985).
This section applies to lactococcal phages only. Generally broth-grown cells plate phages more efficiently than milk-grown cells. For this reason broth derived cultures are usually used to determine phage levels in cheese plants. The explanation for this phenomenon may be the tendency for phage-sensitive variants to accumulate in broth media (Limsowtin et al. 1978). However, milk-grown cells have been found to be more sensitive for some phages (Lawrence, 1978), and this worker has suggested that the use of cells which have been derived from both broth and milk media might reflect more accurately the phage situation in cheese factories.
Additionally, certain phages, often called 'raw milk' phages, inhibit starter activity in pasteurised milk but not in severely heated milk (100°C, 10 min or of higher heat treatment). It seems likely that these viruses require cations that have been complexed because of the severe treatment. Both calcium and magnesium may be involved. These phages can be detected using the plaque assay or by activity tests in pasteurised milk. Such milk must be free from phage and antibiotics and precautions must be taken to avoid culture inhibition due to immunoglobins (agglutinins) by agar or rennet addition.
While many phage have a narrow host range, it is now well established that some phages can attack several or more strains or even other species e.g. phage P100 is a virulent, wide-host-range phage capable of infecting Listeria strains of various species and serotypes, phage Felix O1, lyses almost all Salmonella serotypes while phage S1 lyses a broad range of pseudomonads. Bacteriophages D29 and TM4 are able to infect a wide range of mycobacteria, including pathogenic and non-pathogenic species.
A strain attacked by a phage originally isolated from another strain is designated as a heterologous host for that phage. In some instances the phage may plate with higher efficiency or be easier to enumerate on a heterologous host.
particularly in working with slower growing apparently homologous hosts.
Some phage-host systems will not produce plaques unless host cells are in the correct growth phase.
Logarithmic and stationary phase cells are widely used in phage assay. Maximum EOP will often require cells from a particular phase and is phage-host dependant. Lowrie (1974) reported that late stationary phase Lc. lactis subsp cremoris AM1 was required for the assay of the temperate phage r1t, as logarithmic phase cells gave no plaque formation.
Choice of an appropriate growth medium is essential for plque formation. The medium must support the growth of the host, be free of virus inhibitors and agents that chelate co-factors. The research with lactococcal phages is worth reviewing for those working with other bacterial groups since a lot of the work particularly that concerning buffers can readily be applied to phages of other genera.
Phage preparations must be diluted to obtain a plate with a countable number of plaques. Because some phage preparations can be of particularly high titre and several serial dilution steps may be required, the possibility of phage inactivation during the dilution process should be considered.
Phages can be rapidly in-activated when aqueous solutions are vigorously shaken in air (Adams, 1959). This surface inactivation can be prevented by adding enough protein to the diluent to saturate the gas-liquid interface and prevent access of the virus to the surface; 10 to 100 µg of gelatin per mL of diluent were considered adequate to prevent this problem (Adams, 1959).
A wide range of diluents are used in phage assay and range from buffer solutions with or without soluble protein, salts solutions, peptone solutions to the phage-plating medium (minus the agar component) either un-diluted or at around 50% strength to pure water. Water as a diluent is not recommended and its use as a diluent may reduce the plaque count for some phages.
The effect of the diluents on host growth must also be considered e.g. the growth of Sphaerotilus natans is inhibited by sodium ions hence diluents for phages for this organism should not contain added sodium.
There is little information concerning the effect of diluent on the EOP of lactococcal phages. In general, there appears to be no major problems in using standard 25 % Ringers solution as diluent.
Temperature can have a major effect on phage replication. Some phage-infected bacteria do not produce phage at 28°C but do so at 18°C. Others do not produce phage at temperatures >34°C while certain phages are produced at 37°C but not at lower temperatures. The existence of phages that replicate at all temperatures permitting host growth must also be recognised.
This is particularly true for lactococcal phages. Sanders and Klaenhammer (1984) made the statement "Temperature has been shown to have a dramatic, yet varied, effect on phage-host interactions of the lactic streptococci (now lactococci)". This statement is as valid today as in 1984 and also applies to other phage-host systems.
Recently Carlos G. Leon-Velarde and his colleagues (Leon-Velarde et. al., 2016) have characterised two phages for Yersinia enterocolitica, TG1 and φR1-RT. Both phages replicate at 4°C, 10°C, 16°C and 22°C but not at 37°C and have a receptor on the outer membrane protein (OmpF) of Y. enterocolitica . OmpF is not expressed at 37°C which renders cells grown at this temperature resistant to lysis by these phages.
While a standard temperature of 30°C is often used in studies of phage enumeration the temperatures used in the fermentation process, or in the environment of concern, should be reflected in methods used to detect phage if reliable results, including plaque formation, are to be obtained.
In the double agar method as described by Adams (1959) phage adsorption occurs in molten agar at around 46° to 50°C. In a study of factors affecting the EOP of phages for Lc. lactis subsp cremoris strains AM1 and AM2, the author found that if adsorption of phage to host cells took place at 20°C, prior to agar addition, a significant increase in EOP was obtained (Table 3). Plaque diameter was slightly increased, also. Although this modification did not give a marked increase in EOP for lactococcal phages generally, a slight increase in plaque diameter was often found.
Interestingly Pearce (1978) and Sanders and Klaenhammer (1980) have shown that heating some lactococci from 40° to 50°C before plaque assay markedly increased the EOP. The heat treatments used were considered to reduce the activity of modification-restriction systems.
Incorporation of glycine in growth media
Lillehaug (1996) has found that glycine incorporation in growth media markedly increased plaque size for temperate phages. The concentration required for maximum plaque size was host dependent.
Santos et al. (2009) have summarised work aimed at improving the visibility of hard-to-see plaques, micro plaques. It would appear that the dyes most frequently used to improve contrast between the zone of confluent bacterial growth and the plaque (e.g. 2,3,5-triphenyltetrazolium chloride and 2,5-diphenyl-3 [alphanaphthyl]-tetrazolium chloride) may actually supress phage tires suggesting that they have limited application. They also commented on the limitations of using ferric ammonium citrate and sodium thiosulfate (FACST) to enhance plaque visualization; not only is this approach limited to hosts that produce hydrogen sulphide, plaque counts have to be made within 12 h of plating.
Following several reports e.g. Maiques et al. (2006) indicating that antibiotics particularly beta-lactam antibiotics may increase phage production in infected phage-host systems a number of workers have investigated the potential for improving plaque formation using antibiotics. Santos et al. (2009) studied the effects of ampicillin, penicillin G, kanamycin, rifampicin and tetracycline hydrochloride (and glycerol) on plaque formation by four phages, representing the three families of the Caudovirales. These phages were active against Salmonella enterica Enteritidis, Pseudomonas fluorescens and / or Staphylococcus lentus.
Only penicillin G, ampicillin, cefotaxime and tetracycline were found to increase plaque size. These antibiotics all damage bacterial DNA and induce the SOS cellular repair response in cells. However, antibiotic addition to the top layer was reported to result in plaques of variable size that were not distributed uniformly. This was overcome by the addition of the antibiotic to both layers. A further increase in plaque size was achieved using 5% glycerol, where the authors also reported an increase in plaque contrast.
Typical results are shown in Figure 2.
DLA is control antibiotic and glycerol-free agar. PAMA is modified agar containing antibiotics and glycerol.
Based on these results it is clear that a single optimised antibiotic based protocol is unlikely to work for all phages and that an individual phage-host antibiotic dosage will be required for optimum results.
Workers using beta-lactam antibiotics and fluoroquinolones should be aware of the potential for prophage induction (Maiques et al. 2006; Allen et al., 2011) associated with the SOS response and should include appropriate controls.
It should be clear to readers that there may be many reasons why it might be difficult to get a newly isolated phage to form plaques. However, I have had discussions with people working with well characterised phages e.g. the listerial P100 phage who have had difficulties in getting phages to propagate let alone form plaques. In some cases the phage required divalent cations and the calcium ions added to the media were reacting with medium components and precipitating out of solution. Phage propagation was not possible since the required ionic calcium concentration had not been attained.
Given that it may be necessary to test a range of conditions to find a host for a new phage and to get the phage to form plaques e.g. log and stationary phase cultures, a range of temperatures, low strength agarose top agar overlays, media supplemented with Ca++ and Mg++, media that do not chelate cations it could be argued that with the necessary knowledge, practical aptitude, determination and effort many new phages can be got to form plaques. The significant work sometimes involved in this exercise must be recognised however. And yes despite everything failure is a possibility! Are there other biological approaches?
The possibility of engineering a range of universal hosts e.g. listerial, clostridial, lactococcal etc. strains is worth considering. The advantages of such a host are obvious as far as quality assurance of mixed culture fermentations is concerned.
Phage resistance mechanisms have been extensively studied in recent years and there are over 50 known natural phage resistance mechanisms found in lactococci (Coffey and Ross, 2002) for example. Most natural defensive mechanisms are not integrated into the bacterial chromosome and are on plasmids. However, another resistance mechanism, superinfection immunity is coded on the bacterial chromosome and is conferred by resident prophages.
Bacteria can easily be cured of plasmids, and many prophages, suggesting the possibility of engineering indicator strains with broad phage sensitivity.
Because of the difficulties in isolating and culturing bacteriophages techniques have been developed using molecular biology to study phage populations. Metagenomic methods are well established for characterising bacterial populations (for a review see Green and Keller, 2006). However, phage species generally lack conserved proteins and genes and it took the development of sophisticated sequencing techniques combined with improvements in computational techniques to study the diversity and dynamics of phage populations using metagenomics.
Gene sequences can now be obtained directly from environmental samples and virus types identified without culturing; this is generally referred to as viral shotgun metagenomics. Sophisticated sampling, purification and gene amplification techniques are required along with access to viral metagenomic libraries. There are at least five published metagenomic libraries for double stranded DNA phages (only phages active against lactic acid bacteria to date) but only limited information available for single stranded DNA phages and for RNA-phages. Additionally some major food companies and research laboratories will have gene libraries for some of the phages in their collections.
Quiberoni et al. (2010) have reviewed the use of molecular methods to identify phages for Str. thermophilus. Dr Giuseppe Aprea (Aprea et al., 2015) has used multiplex PCR to screen and study the diversity of phages against lactic acid bacteria in the whey from cheese made with artisanal Water Buffalo Mozzarella starters. The results showed phages belonging to the phage P335 group against Lactococcus lactis and also phages against Lb. delbruekii and Str. thermophilus present in whey from most slow-acid cheeses. The PCR technique could detect 102 PFU/ml of a previously isolated lactococcal phage.
It will be interesting to see how this area develops in the future.
Abedon, S.T. and Yin, J. (2009). Bacteriophage plaques: theory and analysis. Methods Mol Biol. 501, 161-174.
Adams, M.H. (1959) In 'Bacteriophages', Interscience Publishers, Inc., New York.
Allen, H.K., Looft, T., Bayles, D.O., Humphrey, S., Levine, U.Y., Alt, D., Stanton, T.B. (2011). Antibiotics in feed induce prophages in swine fecal microbiomes. MBio.2011;2:e00260-11.
Aprea, G., Mullan, W.M.A., Murru, N., Tozzi, M. and Cortesi, M. L. (2015). Multiplex PCR to detect bacteriophages from natural whey cultures of buffalo milk and characterisation of two phages active against Lactococcus lactis, ɸApr-1 and ɸApr-2. Accepted for publication in Veterinaria Italiana.
Coffey, A. and R.P. Ross. (2002). Bacteriophage resistance systems in dairy starter strains: molecular analysis to application. Lactic Acid Bacteria, Genetics Metabolism and Applications. Antonie van Leeuwenhoek. 82, 303-321.
de Man, J.C., Rogosa, M. and Sharpe, E. (1960) Medium for the cultivation of Lactobacilli. J. Appl. Bacteriol. 23, 103-135.
Douglas, J. (1971) A critical review of the use of glycerophosphates in microbiological media. Lab. Pract. 20, 414-417.
Foddai, A., Elliott,C.T. and Grant, I.R. (2009). Optimization of a phage amplification assay to permit accurate enumeration of viable Mycobacterium avium subsp. paratuberculosis cells. Applied and Environmental Microbiology. 75, 896–3902.
Fullner, K.J. and Hatfull, G.F. (1997). Mycobacteriophage L5 infection of Mycobacterium bovis BCG: Implications for phage genetics in the slow-growing mycobacteria. Mol Microbiol. 26, 755-66.
Green, B. D., and M. Keller. (2006). Capturing the uncultivated majority. Curr. Opin. Biotechnol. 17:236–240.
Heap, H. A., Limsowtin, G. K. Y.and Lawrence, R. C. (1978). Contribution of Streptococcus lactis strains in raw milk to phage infection in commercial cheese factories. New Zealand Journal of Dairy Science and Technology. 13(1): 16-22.
Kropinski, A.M., Mazzocco, A., Waddell, T.E., Lingohr, E., Johnson R.P. (2009). Enumeration of bacteriophages by double agar overlay plaque assay. Methods in Molecular Biology Methods, 501:69-76.
Leon-Velarde et al. (2016) Yersinia enterocolitica specific infection by bacteriophages TG1 and φR1-RT is dependent on temperature regulated expression of the phage host receptor. Accepted for publication in Appl Env Microbiol. Further details will be given when published.
Lawrence, R.C. (1978). Action of bacteriophages on lactic acid bacteria: consequence and protection. N. Z. J. Dairy Sci. Technol. 13:129.
Lillehaug, D. (1997). An improved plaque assay for poor plaque-producing temperate lactococcal bacteriophages. Journal of Applied Microbiology. 83, 85–90.
Lowrie, R.J. (1974) Lysogenic strains of group N lactic streptococci. Appl. Microbiol. 27, 210-217.
Lowrie, R.J. and Pearce, L.E. (1971) The plating efficiency of bacteriophages of lactic streptococci. N.Z. J. Dairy Sci. Technol. 6, 166-171.
Maiques, E., Ubeda, C., Campoy, S., Salvador, N., Lasa I., Novick, R.P., Barbé, J., Penadés, J.R. (2006). Beta-lactam antibiotics induce the SOS response and horizontal transfer of virulence factors in Staphylococcus aureus. J. Bacteriol. 188, 2726–2729.
Mullan, W.M.A. (1979). Lactic streptococcal bacteriophage enumeration. Dairy Industries. 44, (7):1-15.
Mullan, Michael (2012). Bacteriophage and food fermentations. Phage assay and enumeration. In: L Laboratori Nazionali di Riferimento per Listeria monocytogenes e Camplpylobacter. Seminario. 10-12 December, 2012, Teramo, Italy. This can be downloaded from www.researchgate.net .
McConnell, M. and A. Wright. (1975). An anaerobic technique for increasing bacteriophage plaque size. Virology 65:588–590.
Paranchych,W. (1966) Stages in phage R17 infection: The role of divalent cations. Virology. 28, 90-99.
Pearce, L. E. (1978). The effect of host-controlled modification on the replication rate of a lactic streptococcal bacteriophage. N.Z.J. Dairy Sci.Technol. 13, 166-171.
Qanber, A.A. and J. Douglas. (1976). Enhancement of plaque size of a staphylococcal phage. Journal of Applied Bacteriology. 40:109–110.
Quiberoni, A., Moineau, S., Rousseau, G.M., Reinheimer, J. and Ackermann, H-W.(2010). Streptococcus thermophilus bacteriophages. International Dairy Journal. 20, 657-664.
Sanders, M. E. and Klaenhammer, T. R. (1980). Restriction and modification in group N. streptococci: effect of heat on development of modified lytic bacteriophage. Appl. Environ. Microbiol. 40, 500-506.
Sanders, M. E. and Klaenhammer, T. R. (1984). Phage resistance in a phage-insensitive strain of Streptococcus lactis: temperature-dependent phage development and host-controlled phage replication. Appl. Environ. Microbiol. 47:979-985.
Santos, S.B., Carvalho, C.M., Sillankorva, S., Nicolau, A., Ferreira, E.C., and Azeredo, J. (2009). The use of antibiotics to improve phage detection and enumeration by the double-layer agar technique. BMC Microbiol. 9, 10.
Serwer, P., Hayes, S.J., Thomas, J.A., Hardies, S.C. (2007). Propagating the missing bacteriophages: A large bacteriophage in a new class. Virol. J. 4, 21.
Terzaghi, Betty E. and Sandine, W.E. (1975) Improved medium for lactic streptococci and their bacteriophages. Appl. Microbiol. 29, 807-813.
Turner, G.E. (1948) Dynamics of the Streptococcus lactis bacteriophage relationship. Ph.D. Thesis. Iowa State University.
Williamson, K.E., Radosevich M., Wommack K.E. (2005). Abundance and diversity of viruses in six Delaware soils. Appl Environ Microbiol. 71, 3119-3125.
Zameni, E., Mullan, W.M.A. and Espie, W.E. (1985). Lactic streptococcal bacteriophage enumeration. Observations on the influence of host cell concentration on plaque formation. Irish Journal of Food Science and Technology. 9:79.