Interest in the use of feed enzymes in corn soy diets has arisen as a result of dramatic increases in the shadow prices of energy and phosphorus in particular, but also to a lesser extent that of amino acids. Whilst these values have abated, the interest in such products has remained.

With regards to non-phytase enzymes, xylanases, ß 1-4 glucanases, mannanases and, to some extent, amylases, have constituted the majority of enzyme classes employed in corn soy based diets, yet their benefits and limitations are only beginning to be realised.

Recently, however, there has been a surge in offerings based significantly, or even entirely, upon proteases. Some of the issues that need to be considered when using a protease, and are relevant for most enzymes, are examined below.

Is the data based on in vitro studies?

In vitro studies are not always an effective means of evaluating feed enzymes. In the case of a protease, any evaluation must take into account that the proventriculus/gizzard and the small intestine of the bird already produce significant quantities of pepsin and pancreatic enzymes, respectively. It is no surprise that feeding a protease would release peptides from dietary proteins, but does it synergize with the pepsin and pancreatic enzymes present in the gut?

A significant issue is whether the simulation of the intestine is conducted using relevant conditions. An assay deliberately short in time, and employing a pH and temperature that favour the exogenous rather than the endogenous protease is not appropriate and likely to give overly optimistic results. Similarly, use of low dosages of pepsin and/or trypsin/chymotrypsin will not reflect the value of the exogenous protease in vivo.

What are the targeted protein ingredients, and what kind of protease is being offered?

If the majority of the trials were conducted in diets containing corn and soy as the sole protein sources, does the product work when meat and bone, or other animal by-products are used? The specificity of the protease may reduce its utility in such diets. Also, use of extremely low protein diets may not give results that are reproducible in commercial diets.

What is the nature of the protease in the product? There are several families of protease available and all have different characteristics and substrate specificity. Firstly, proteases can be divided into two broad categories, exo- and endo-acting. Proteinases or endopeptidases cleave inside the protein to produce large peptides. Peptidases or exopeptidases cleave the end fragments of proteins to produce small peptides and amino acids.

Understanding the nature of the protease is important as some proteases are endogenously produced and others are not. A protease with similar characteristics to endogenous proteases may have short-lived effects only in the neonate because the animal will produce enough.

What proportion of the in vivo data shows performance benefits to 42 days of age?

The most positive responses to proteases are in young animals when endogenous protease production may be limited. Digestibility and performance data generated with 21 day old birds will likely not represent the effect out to 42 days, since endogenous protease production is sufficient in older birds. In the data presented in Table 1 there is an effect of the proteases in addition to the xylanase and amylase on FCR to 21 days, but this is lost at 42 days of age (date not shown).

Also, note that the xylanase and amylase alone give a benefit to 42 days in terms of gain but there is no effect of protease, even at 21 days of age, when added on top of such an enzyme.

The protein concentration in monogastric diets generally decreases with an animal’s age, whereas the ability to produce proteases increases. Consequently, the requirement for proteolytic intervention is less acute in the older animal compared with the neonate.

Is the protease fed in a diet where ordinarily another enzyme would be used – e.g. a phytase or a xylanase, and is the protease truly pure?

If another enzyme would ordinarily be fed with the protease, then it is prudent to check that the protease does not degrade the other enzyme. Most protease work is conducted in low protein diets that are devoid of other added enzymes. This is essential if the compounder is not to lose efficacy of the “other” enzymes which are assumed to deliver efficacy in the animal.

The very fact that much of the work on protease products is conducted in isolation means that additivity with other enzymes is unknown. It seems likely, for example, that the value of an NSP’ase is reduced when used with a phytase in corn soy diets. It is likely that the value of a protease is similarly diminished in the presence of other enzymes and vice versa.

Is the enzyme really a protease? Quite often crude fermentation products or products from host organisms which produce other enzyme activities are presented as a protease. It may well be that any response observed is related as much to the contaminants as it is the declared activity.

Are many trials presented, and if so, are any negative?

The value of an enzyme in the commercial world cannot be predicted from one or two trials. So many factors interact with the enzyme to determine the scale of the response that it is important to test the enzyme under a variety of circumstances.

In all large data sets, there are always negative responses. In fact, even with a product that works every time, it is likely, by random chance, that at least one trial would be negative. Experience suggests that the very best of corn soy enzymes has a strike rate of approximately 80%. Anything more optimistic questions whether some trials have been ignored.

Is it heat stable? If so, is it coated? How much of the data generated is produced in pelleted diets?

Many of the benefits of a protease in a mash diet may be absent in a pelleted diet. There are several possible reasons for this;

  • The protease may have poor heat stability. Losses during pelleting may reduce efficacy.
  • If the targets of the protease are heat labile anti-nutrients or proteins which become sufficiently denatured during pelleting to allow normal digestion.
  • If the protease is coated, is it released rapidly enough in vivo to deliver the responses reported in the literature? These reported responses are often determined using a non-coated product in mash or low temperature pelleted diets.