The basic principles of phage control in commercial plants have been known since the early 1940s and the pioneering work of Dr Hugh Whitehead and his colleagues in New Zealand. The review by Whitehead and Hunter (1945)* on the measures that were being used in New Zealand to control slow acid production due to phage infection is still of relevance to factory managers today.

The 1945 review focused on whey as the vehicle for phage transmission, and on work designed to break the cycle of phage infection. Even in 1945 they recognised the challenge of keeping phage concentrations low in the environment, the need for special facilities to produce phage-free bulk starter in a potentially phage-infected environment, the possibility of raw milk being contaminated with phage because the cans, tankers today, carrying whey, and also used to carry raw milk, could contaminate raw milk. They were also aware that whey separators and 'splashes' of whey produced aerosols and that these would enable phage to become airborne.

Off course they were not aware of lysogency and the possibility of phage arising from the starter or lactococci in raw milk.

Modern methods for controlling phage problems in cheesemaking include:-

1. Use of starters containing phage-unrelated or phage-insensitive strains.

2. Production of 'phage-free' bulk starter
 • Aseptic propagation systems
 • Use of phage-inhibitory media
 • Segregation of starter room and cheese process equipment CIP-systems
 • Removal of deposits on bulk starter vessels
 • Ensuring the ''head space' is minimised in bulk starter vessels and sterilised.

3. Minimising the concentration of phage in processing plants
 • Culture rotations
 • Air conditioning
 • Aerosol generation/fumigation
 • Cleaning/chlorination of vats between refills
 • Good factory design. Location of whey storage tanks, whey handling systems.

3. Other measures
 • Use of DVI/DVS cultures
 • Inspection of jackets/agitators for pin holing
 • Early renneting.

Use of starters containing phage-unrelated or phage-insensitive strains.

In general the large scale commercial manufacture of Cheddar and other hard cheeses in Europe, North America, Australia and New Zealand is achieved using defined strain cultures. NOTE! This does not mean that good cheese cannot be produced using undefined mixed strain cultures. Defined strain cultures generally contain two or more phage- unrelated strains of lactococci although strains of Str. thermophilus may also be used. Strains are chosen using physiological and phage sensitivity criteria. Cultures may be used with or without rotations. Some of these cultures are now used continuously without rotations for long periods; however, good phage monitoring procedures must be in place.

New sources of strains, in particular from artisanal cultures as used in France and Italy, continue to be sought and phage-insensitive variants arising naturally from infected cultures are being continually evaluated for commercial use.

Production of phage-free bulk starter.

When Whitehead and Cox (1935) discovered that lactococcal phages caused starter cell lysis, they recognised that the first object of any starter control system had to be the protection of the bulk starter and/or mother culture from phage infection.

The latent period and burst size characteristics of the phage-host interaction of lactococci are such that a very small number of phage particles entering the bulk starter milk at the time of inoculation will either result in the complete lysis of the host before the bulk starter coagulates, or, if the culture coagulates, lysis will occur during cheesemaking. It is these latter, apparently normal starters, which are responsible for much of the economic loss from 'dead' vats .Following the demonstration by Whitehead and Hunter (1941) of the importance of airborne phage, considerable effort has been devoted to preventing airborne phage contamination of the heat-treated starter milk.

Aseptic techniques for the propagation of starters have now reached a high level of sophistication. Generally these are variants of early methods, developed by Jones (1956) and Lewis (1956). The Jones method involved the production of bulk starter in a vessel sealed with water which can be pressurised with sterile air when the heat treated milk is cooling. Initially, difficulties were experienced in obtaining a source of sterile air but these were finally solved by the use of high efficiency particulate (HEPA) filters. A major weakness of the Jones system was susceptibility to phage during inoculation and in some factories inoculating the bulk starter through a flaming ring was common. I will deal in more detail with bulk starter production elsewhere.

Lewis (1956) developed a system for inoculating laboratory and intermediate cultures, without the risk of phage contamination, in which syringes were used to transfer cultures through self-sealing rubber seals covered with hypochlorite solution. This method is still used to ensure aseptic inoculation of bulk starter vessels.

Additional measures including maintaining the starter room under positive pressure with sterile air to prevent the entry of phage, and use of aerosol disinfection of the air in the laboratory and cheese room with hypochlorite and other anti-phage agents have also been used to reducing the concentration of airborne phage and to minimise the probability of phage infection during the cooling of bulk starter milk and during inoculation. Discussion of fogging agents will be included elsewhere.

Location of the bulk starter facility is also critical; it should be located in a 'sealed unit' distant from devices and practices producing phage aerosols e.g. whey separators and access should be limited to personnel who have not been in contact with whey and raw milk. Cleaning and sterilisation of tanks, pipe lines and valves are very important. Ideally a dedicated CIP system isolated and distinct from the cheese system should be used. Periodically tanks and pipelines should be dismantled to ensure that there are no deposits and treatment with dilute phosphoric acid or appropriate acidulant should be used to remove deposits. More information is available here. 

Bacteriophage inhibitory media.

Most, but not all, lactococcal phages require divalent cations in particular calcium for multiplication. Protection of the bulk starter against phage contamination can also be achieved either by removing the calcium from the starter milk by ion exchange or by sequestering it chemically. There are a number of commercial phage inhibitory media available for bulk starter production. Some of these media can also be combined with external pH control to produce starter with about 10 times the cell density of 'normal starter'. The high cell density is also helpful in reducing the probability of phage problems in the cheese vat, since more cells can be sacrificed without apparent acidification problems.

These media do have some limitations in that that all strains will grow well and as these media can only be used for preparing phage-free bulk starter, it is not surprising that they have not entirely eliminated starter problems.

There were early suggestions that the continuous use of phage resistant media could lead to the selection of these phages that do not require cations but as yet, there is no evidence to suggest that they are a problem commercially.

Culture rotations.

Although the production of phage-free bulk starter or the use of phage-free starter concentrate is perhaps the major defense against phage induced slowness in the cheese vat, the starter is still susceptible to phage attack during cheese manufacture. Obviously relatively high concentrations of phage must be present in the cheese milk for this to occur.

Cheese whey which may contain very high titres of phage is regarded as the primary source of phage infection during cheese manufacture. Whey may come into contact with cheesemaking equipment directly, or by splashing, or may be distributed through the air by the atomising action of whey separators. If the whey contains very high levels of phage i.e. 109-1010 PFU/ml then there may be sufficient phage in the air of the creamery to contaminate the cheese milk and utensils. If cheese were made with the same starter next day, then slow acid production or even failure of the fermentation could occur. Anderson and Meanwell (1942) first suggested the rotation of phage unrelated starters. This is now a well established practice for controlling the build up of specific phages in cheese factories. Two objectives are attained by this method:-

1. The probability of infection of the bulk starter with its specific phage is reduced because, after three to four days, the bulk of the airborne phage will have settled out.
2. The probability of the starter in the cheese-milk being affected by the airborne phage which may fall into the cheese vat and contaminate the cheese milk and cheesemaking equipment is also reduced because, after three to four days, the bulk of the airborne phage will have settled out.

The importance of not introducing starters into rotations, where the cross phage relationships have not been determined, has been known for more than 30 years, and should not be ignored. Basically the new starter could provide a link allowing a faster developing phage to attack an existing starter.

With properly constituted modern defined cultures there are plants who do not use rotations. These, however, are in the minority.

Elimination of the ripening period.

The addition of rennet, practically simultaneously with phage-free bulk starter or starter concentrate can protect starter in vat milk from phage attack. This technique, 'early renneting', originated from the observation by Hunter (1944) that renneted milk could support the growth of starter in spite of phage infection. The protective effect is largely mechanical; the starter cells are 'trapped' or immobilised in a protective gel matrix. This protective gel impedes the movement of phage and dramatically slows down the process of phage infection and potentially enables virtually normal cheese manufacture to take place. With some phage-host systems there is some evidence that rennet may also reduce or interfere with the process of phage adsorption. Interestingly more recent work has shown that some artisanal rennets may be a source of phage (Mahony et al. 2017).

Direct inoculation of the cheese milk with starter concentrates.

The production of phage-free bulk starter requires both special equipment and even more critically, appropriately trained and skilled staff.

The addition of concentrated starter cultures to the cheese-milk, avoiding both the mother culture and bulk culture stages is an important control mechanism for those factories that do not have the resource to produce phage-free bulk starter.

Starter concentrates are available in two forms, freeze dried or deep-frozen pellets.They are expensive compared with bulk starter. Starter concentrates are discussed in more detail elsewhere. Starter concentrates are generally referred to as Direct Vat Inoculation (DVI) or Direct Vat Set (DVS) cultures.

Over the last 5-10 years starter suppliers have started including Str. thermophilus in concentrates for Cheddar manufacture. The main reason for doing this was to reduce their production costs and to maintain competitive culture supply costs. While initially there was also some phage advantages, phages active against Str. thermophilus have now appeared and have caused slow-acid production problems.

Good factory design.

This is a very important area and includes the correct people and process flows; the correct location of key items of plant; the correct physical separation between raw milk, whey and starter streams; the use of dedicated heat exchangers and fat-separators for each stream as required and separate CIP systems.

Cleaning/chlorination and disinfection of vats between refills

In modern cheesemaking plants, vats can be filled and refilled many times a day. With this practice there is potential for significant phage build up.

While this can be reduced by using rotations and by rinsing with water followed by a chlorine rinse, use of chlorine compounds has the potential to generate unwanted disinfection byproducts.  European companies that use the cheese whey for further processing are increasingly just rising with water between fills  and relying on rotations to reduce phage build-up or replacing hypochlorite with peracetic acid.


Practical phage control will use a range of preventive measures, this is analogous to hurdle technology used in food microbiology. Over reliance on any one control technique should be avoided. A good phage control system will be integrated with an effective phage monitoring system. For the latter to be successful, an effective partnership with a starter supplier will be required.

*Whitehead, H. R. and Hunter, G.J.E. (1945) Bacteriophage infection in cheese manufacture. J. Dairy Res. 14, 64-80.

How to cite this article

Mullan, W.M.A. (2003). [On-line]. Available from: . Accessed: 25 April, 2024. Updated November 2017; April 2021.