The Enzyme: Phytase

Notionally, phytases have the capacity to degrade IP6 phytate completely to
inositol and to liberate six P moieties. However, because the P moiety axially
located at C2 is not readily released, complete dephosphorylation of phytate by
phytase probably does not occur in pigs and poultry. By contrast, there is a
possibility that endogenous phosphatases (associated with the brush border) do
provide some inositol, particularly in the more distal regions of the small
intestine. Thus the role of inositol genesis by microbial phytase and phosphatases
in the overall effi cacy of such products is obscure and warrants further study.
Phytases and phosphatases exist widely in nature, but four sources of phytase
activity are relevant in target species.
Sources of phytase activity
Intrinsic ‘plant’ phytase
Certain feed ingredients, particularly wheat and its by-products (Peers, 1953),
possess intrinsic phytase activity. However, the importance of plant phytase in
standard diets is questionable because it is less effective than microbial phytases
at gastrointestinal pH and may be inactivated by acidic pH levels in the gut.
Moreover, the practical importance of plant phytase is diminished because it
will be reduced or even eliminated by steam-pelleting of pig and poultry diets.
Plant phytases are heat labile and, in purifi ed form, most are destroyed at
temperatures above 70°C within minutes (Konietzny and Greiner, 2002). As
reported by Jongbloed and Kemme (1990), steam-pelleting a diet based on
wheat, maize and soybean meal at 80°C eliminated wheat phytase activity and
reduced total-tract P digestibility by 37% in pigs. It follows that responses to
microbial phytases may be compromised by the presence of plant phytase
activity in the diet, so wheat may be ‘pre-pelleted’ in feeding studies to eliminate
intrinsic phytase and avoid this potential confounding factor. However, robust
responses to microbial phytase have been reported despite the dietary presence
of wheat phytase in weaner pigs (Campbell et al., 1995). This suggests that
the presence of plant phytase may not necessarily compromise responses to
microbial phytases. Indeed, Zimmermann et al. (2002) reported that in vivo
effi cacy of plant-derived phytases was only 40% of microbial phytase on a unit
for unit basis, suggesting that plant-derived phytases do not possess
characteristics optimal for effi cacy in the gastrointestinal tract.
Endogenous mucosal phytase
Patwardhan (1937) fi rst detected the presence of mucosal phytase activity in
rats, and it has been identifi ed in the small intestine of pigs (Hu et al., 1996)
and poultry (Maenz and Classen, 1998). Nevertheless, the importance of
mucosal phytase is usually dismissed, but its activity may be governed by dietary
non-phytate P levels. However, dietary Ca levels appear critical, as Tamim et
al. (2004) reported an ileal degradation coeffi cient for phytate of 0.692 in
maize–soy broiler diets containing 2.8 g phytate-P kg–1 at a dietary Ca level of
only 2.0 g kg–1. However, when Ca was increased to 7.0 g kg–1, the coeffi cient
was noticeably reduced to 0.254. Clearly, Ca has a substantial impact on the
effi cacy of mucosal phytase and, presumably, this is largely a consequence of
the formation of insoluble Ca–phytate complexes at pH approaching neutrality
in the small intestine (Wise, 1983). Consequently, the extent of phytate
degradation generated by mucosal phytase will be limited by the Ca levels in
practical pig and poultry diets.
Gut microfl oral phytase
The microfl oral population in the gastrointestinal tract, especially in the hindgut,
is known to generate phytase activity, although degradation of phytate in the
hind gut is of relatively little importance. It is, however, often assumed that
undigested phytate-P is excreted by pigs and poultry, but the amount may be
markedly reduced by hindgut fermentation, particularly in pigs. While hindgut
fermentation of phytate-P may be of value to coprophagic animals, this
confounds total-tract assessments of phytate degradation since P released postileum
appears to be of little value to the animal (Zimmerman et al. 2002).
Exogenous microbial phytase (feed enzymes)
Presently, the majority of phytases are derived from fungi (e.g. A. niger) or,
more recently, bacteria (e.g. Escherichia coli), and the fermentative production
processes depend on genetically modifi ed organisms. However, it should be
noted that genetically modifi ed material is not found in preparations of phytase
feed enzymes. It is also probable that the purity, or the lack of enzymic sideactivities,
and the yield of phytase activity have increased over time with the
refi nement of production processes. A corollary of this development is the
possibility that the microbial phytases evaluated in early studies (Simons at al.,
1990; Beers and Jongbloed, 1992; Ketaren et al., 1993) are not identical to
the feed enzymes presently available.
Enzymatic dephosphorylation of phytate in pigs and poultry
The main sites of phytate degradation by microbial phytases are the stomach
in pigs and the forestomach (crop, proventriculus and gizzard) in poultry, with
relatively little degradation in the distal gastrointestinal tract. The extent and
rapidity of dephosphorylation is critical to both the destruction of phytate (and
so removal of the associated anti-nutritive effect) and the P equivalence of
phytase. Arguably, the P equivalence of phytase is a simple function of dietary
phytate levels and the degree to which it is hydrolysed. Equally, the amelioration
of the anti-nutritive properties of phytate should be governed by the extent and
timing of its degradation.
However, surprisingly few studies have investigated the dephosphorylation
of phytate along the gastrointestinal tract. Taken together, two broiler studies
suggest that degradation of phytate by 500 FTU A. niger phytase kg–1 does
not exceed 35% at the level of the ileum (Camden et al., 2001; Tamim et al.,
2004). In layers, van der Klis et al. (1997) reported that 500 FTU A. niger
phytase kg–1 increased ileal degradation of phytate (0.661 versus 0.081),
which indicates that microbial phytase degraded 58% of dietary phytate. This
comparison suggests that phytase is more effective in laying hens than in
broiler chickens, which may be due to longer digesta retention times in the
forestomach and is refl ected in lower recommended phytase inclusion rates for
layer diets than broiler diets.
More relevant studies have been completed in pigs (Jongbloed et al.,
1992; Mroz et al., 1994; Rapp et al., 2001; Kemme et al., 2006); collectively,
these reports indicate that in the order of 50% of dietary phytate is degraded
by microbial phytase at the ileal level. Assuming all dietary phytate is present
as IP6, the uniform hydrolysis of all IP6 to IP3, with the release of three P
moieties per IP6, would correspond to 50% phytate P destruction. However,
dephosphorylation of phytate is a stepwise reaction and a considerable
proportion of undegraded phytate remains intact as IP6 at the ileal level. For
example, in the study of Rapp et al. (2001), 60% of phytate remained intact
as IP6. Thus the following equation illustrates phytate degradation by microbial
phytase where, at the level of the ileum, 50% of phytate-P has been liberated
and the balance is present as either IP6 or a range of lesser myo-inositol
phosphate esters:
IP6 (100% P) ⇒ IP6 (30%) + [IP3, IP2, IP1] (20%) + inorganic P (50%).
The likelihood is that little, if any, IP6 is completely dephosphorylated to
inorganic P and inositol, essentially because microbial phytases do not release
P located at the axial C2 position of the myo-inositol ring. Alternatively, if the
majority of undegraded phytate remains intact as IP6, then this has
consequences. This is because the chelating capacity and anti-nutritive
properties of phytate are disproportionately diminished as IP6 is degraded to
lesser myo-inositol esters, which are relatively innocuous (Luttrell, 1993).
Recent advances in the understanding of the stepwise dephosphorylation of
IP6 have indicated that there is a considerable difference between 6-phytases
and 3-phytases in this regard (Wyss et al. 1999; Greiner et al., 2000, 2001).
While the commercially employed 3-phytases effectively tend to continue their
attack on a selected IP6 molecule until it is reduced to IP1, the 6-phytases seem
to halt their assault, momentarily, on IP4 and lower esters (due to an apparently
higher KM for these substrates). As a result, for the provision of similar quantities
of P as determined by FTU assay, there is considerably more destruction of IP6
by a 6- compared with a 3-phytase. This will clearly infl uence the relative extraphosphoric
effects of the 3- versus the 6-phytases.
Phosphorus and Calcium Equivalence of Phytase
Effectively, phytase is a source of P and Ca following the enzymatic degradation
of phytate and the liberation of P inherent in the substrate and Ca bound to the
phytate. In the formulation of phytase-supplemented diets, P and Ca levels are
usually reduced to accommodate this release of macro-minerals on the basis of
P and Ca equivalency values for phytase. This adjustment in dietary P levels
contributes to the reduction in P excretion, which is a prime objective of
phytase supplementation.
Hoppe and Schwarz (1993) concluded that 500 FTU phytase was
equivalent to 1 g P as monocalcium phosphate in maize–soy pig diets and,
essentially, this precedent remains accepted. For example, the recommendation
of the relevant manufacturer is that 500 FTU A. niger phytase kg–1 is equivalent
to 1.15 g P kg–1 and 1.00 g Ca kg–1 in diets for pigs and broiler chickens, and
broadly similar recommendations are made by other manufacturers.
Interestingly, that phytase liberates somewhat less Ca than P is a concept that
is still accepted, although it may be questioned from a theoretical viewpoint.
Phosphorus equivalence
Up to this point, P equivalence values have been established by incorporating
graded quantities of either an inorganic P source or microbial phytase into a
P-defi cient basal diet. P replacement values are calculated from regression
equations that best describe responses in selected parameters generated by additional P and microbial phytase; usually, the parameters are weight gain
and a measurement of bone mineralization.
In broiler chickens, body weight gain and percentage toe ash are sensitive
indicators of P availability (Potter, 1988) and are usually selected as the
response criteria in P equivalence studies. Selle and Ravindran (2007) reviewed
nine P equivalency studies in broilers in which P-defi cient basal diets contained
an average of 2.00 g non-phytate P kg–1 and 2.37 g phytate-P kg–1, with a
Ca:P ratio of 1.84:1.00. Collectively these studies indicated that approximately
766 FTU phytase kg–1 is equivalent to 1.0 g P kg–1 in broilers, which implies
42% phytate degradation. This phytase equivalency value is less than standard
recommendations, which may indicate that commercial diets contain a surplus
of P as a safeguard, which is arguably the case. However, the accuracy and
relevance of P equivalency studies are questionable because the basal diet, by
defi nition, contains inadequate levels of non-phytate P. As a consequence the
Ca:P ratios may be higher than in standard diets. This could negatively
infl uence the extent of phytate degradation and the P equivalency value.
Alternatively, phytase may positively infl uence weight gain quite independently
of phytate-bound P release, which would tend to infl ate P equivalency values
(Wu et al., 2004).
Reservations in relation to P equivalency studies have been expressed by
other workers (Angel et al., 2002; Driver et al., 2005). However, it appears
that nutritionists are electing to use higher phytase inclusion rates in practice,
which would be expected to increase P equivalency values and permit greater
reductions in dietary P levels and, in turn, amounts of P excreted. This
emphasizes the need to develop more accurate P equivalence values based on
the extent of phytate degradation induced by phytase, coupled with established
dietary phytate concentrations in preference to values derived from ‘classic’ P
equivalency studies.
From basic principles, if phytase degrades 40% of phytate in a broiler diet
containing 2.8 g phytate-P kg–1, then there is a generation of 1.12 g P kg–1.
The P equivalency value of phytase is clearly a function of the dietary substrate
level and the extent of phytate degradation, which are both variables. It is also
clear that the susceptibility to or availability of phytate to phytase hydrolysis
may be ingredient dependent. Leske and Coon (1999) demonstrated that,
although canola meal contained almost twice as much phytate-P as soybean
meal, the subsequent P equivalency of 600 FTU of an Aspergillus phytase was
three times higher in soybean meal compared with canola. Ideally, therefore, P
equivalence values should be based on determined dietary phytate concentrations
and a prediction of phytase-induced substrate degradation. The development
of such an approach should be a future objective to permit more appropriate
manipulations to dietary formulations.
Calcium equivalence
Calcium equivalency studies follow the same principle, where graded levels of
Ca as limestone or microbial phytase are added to Ca-defi cient basal diets. Few Ca equivalency studies have been completed, but it is accepted that for an
Aspergillus phytase 500 FTU kg–1 is equivalent to about 1.00 g Ca kg–1, and
the formulation of phytase-supplemented diets is usually adjusted accordingly.
In early studies, Schöner et al. (1994) reported that a Ca equivalency of 500
FTU phytase kg–1 was approximately 0.44 g Ca kg–1 in broilers, and Kornegay
et al. (1996) found that a Ca equivalency of 500 FTU phytase kg–1 ranged
from 0.38 to 1.08 g Ca kg–1 in pigs. Further confl icting results have been
recorded. Augspurger and Baker (2004) reported that 500 FTU E. coli phytase
kg–1 released 0.90 g Ca kg–1 on the basis of tibia ash in maize-soy broiler diets;
however, Mitchell and Edwards (1996) and Yan et al. (2006) concluded that
phytase had little impact on Ca release in broilers. Ca levels in the basal diet of
equivalency studies are intentionally low. However, Farkvam et al. (1989)
found that increasing dietary Ca concentrations in broiler diets increased the
amount of Ca bound by phytate. Therefore, it follows that Ca-defi cient basal
diets reduce the amount of Ca bound by phytate, which may explain the
inconsistent results and generally low values recorded in calcium equivalency
studies.
The likelihood is that the Ca equivalence of phytase is governed by the
extent of de novo Ca–phytate complex formation in the small intestine. One
phytate molecule may bind up to fi ve Ca atoms as Ca5-K2-phytate; if so, in a
diet containing both phytate (IP6) and Ca at 10 g kg–1, phytate would have the
capacity to bind 3.0 g Ca kg–1 or approximately one-third of dietary Ca.
Simplistically, this suggests that phytase has an equivalency value of 1.5 g Ca
kg–1 assuming a 50% degradation of phytate. However, the capacity of phytate
to complex Ca declines at a disproportionately greater rate as IP6 phytate is
degraded into lesser inositol phosphate esters. Indeed, Luttrell (1993) found
the in vitro Ca-binding affi nity of IP4 to be 32% in comparison with that of IP6,
and the Ca-binding affi nities of IP3, IP2 and IP1 were negligible.
Consequently, it seems likely that, rather than being in parallel, the
liberation of phytate-bound P and Ca by phytase is ‘uncoupled’. The liberation
of P is directly proportional to the extent of phytate degradation, but the
liberation of Ca may exceed this rate and it may be that the enzymatic hydrolysis
of dietary phytate by microbial phytase generates more Ca than P in the initial
phase. This would be particularly so for 6-phytases, which seem to prefer IP6,
IP5 and IP4 as substrates over IP3 and IP2, in contrast to 3-phytases which
seem to have equal affi nity for all. If so, this is not refl ected in matrix values
applied to phytase-supplemented diets.
Calcium is a critical nutrient but, as discussed later, relatively high Ca levels
in pig and poultry diets, particularly as limestone, may have a negative infl uence
on phytase effi cacy. Consequently, Ca levels in phytase-supplemented diets
should be kept to an acceptable minimum and, for this reason alone, more
accurate assessments of the Ca equivalency of phytase are required. It is even
likely that larger, more appropriate, Ca reductions in phytase-supplemented
diets will enhance the effi cacy of the feed enzyme. Release of phosphorus and calcium
Several experiments have been completed where increasing phytase inclusion
rates, at times to apparently high levels, have been evaluated in pigs and
poultry. In several studies where the highest inclusion rate employed was not
excessive (e.g. less than 2500 FTU kg–1), it is not unusual that responses to
phytase observed appeared to plateau or even decline at higher inclusions
(Ravindran et al., 2001). However, in studies where ‘mega-doses’ of phytases
have been investigated (Rosen, 2001; Veum et al. 2006), the data clearly
indicate a log-linear relationship between dose and response, suggesting that
much of the research is conducted at well below the ‘optimum’ inclusion rate
of this enzyme. As a result, the apparent 750–1000 FTU kg–1 optima
determined in studies where dosages do not exceed 2500 FTU kg–1 are often
an artifact of the design of the study.
Of interest is that Nelson et al. (1980) altered the cation–anion balance of
a maize–soy broiler diet with P as mono–dicalcium phosphate and Ca as
limestone. They reported that net increases in cation levels were negatively
correlated with nitrogen-corrected AME (r = –0.72; P <0.01) and digestibility of 17 amino acids (r = –0.79; P <0.01). Microbial phytase induces the release of P and Ca with the potential to impact on the effective cation–anion balance. If, in fact, phytase liberates more Ca than P this would generate a net increase in dietary cationic levels, which would be detrimental on the basis of the Nelson et al. (1980) study. This point is made because, fundamentally, the greatest impact following the dietary inclusion of phytase is on P and Ca availability. Arguably, the consequences of effectively increasing the dietary levels and altering the balance of these two macro-minerals have not received proper consideration, nor have dietary formulations been appropriately adjusted. Therefore, assessments of P and Ca phytase equivalence values at both standard and elevated inclusion levels merit more accurate defi nitions. ‘Protein Effect’ of Phytate and Phytase Offi cer and Batterham (1992a) were probably the fi rst to suggest that microbial phytase has a ‘protein effect’. In grower pigs offered diets based on linola meal as the only protein source, phytase signifi cantly increased the ileal digestibility of nitrogen (22.6%) and lysine (20.3%), and it was suggested that these responses may be ‘due to the release of amino acids bound in phytate linkages’. It is established that phytate can bind protein to form protein–phytate complexes (Cosgrove, 1966; Anderson, 1985), and it follows that the prior hydrolysis of phytate by phytase in the gut would reduce the extent of de novo protein–phytate complex formation (Selle et al., 2000). Phytate is capable of binding up to ten times its weight of protein under in vitro conditions (Kies et al., 2006), which implies that in a diet with 10 g phytate kg–1 and 200 g protein kg–1, half the protein present may be complexed by phytate. It could be anticipated that phytase enhances ileal digestibility of amino acids in pigs and poultry via reductions in protein–phytate complex formation. However, the outcomes of phytase amino acid digestibility assays are inconsistent, particularly in pigs where responses to phytase have often been negligible. Indeed, Adeola and Sands (2003) were inclined to the view that phytase does not have a positive effect on protein utilization in pigs. Nevertheless, regardless of the confl icting data arising from phytase amino acid digestibility studies, some practical nutritionists elect to confer amino acid matrix values to microbial phytase in pig and poultry diet formulations. The protein effect of phytate and phytase is still an open question and, given that microbial phytases have been commercial entities for nearly two decades, it is not acceptable that this fundamental issue remains unresolved. Microbial phytase amino acid digestibility assays in broilers Despite its recognized limitations as a dietary marker (Jagger et al., 1992), chromic oxide has been used in the majority of phytase amino acid digestibility assays. However, in broiler chickens, amino acid digestibility responses to phytase using acid-insoluble ash or titanium oxide have been consistently more pronounced than those involving chromic oxide (Selle et al., 2006; Selle and Ravindran, 2007). It is recognized that ileal digestibility of amino acids is more meaningful than assessments made on a total-tract basis (Ravindran et al., 1999b). Nevertheless, Hassanabadi et al. (2008a,b) determined the infl uence of microbial phytase on total-tract digestibility of amino acids by quantitative excreta collection, which did not involve a dietary marker. Aspergillus niger phytase (500 FTU kg–1) increased mean AID coeffi cients of 13 amino acids by 5.1% (0.902 versus 0.858) in female chicks and by 4.2% (0.889 versus 0.853) in male chicks. The magnitude of these responses is very similar to ileal digestibility assays in which acid-insoluble ash or titanium oxide were used as markers. Eight assays are identifi ed in which the effect of microbial phytase on AID of amino acids was determined in broilers with either acid-insoluble ash or titanium oxide as dietary marker (Table 7.2). In the eight studies, phytase 0.824, over 123 observations (Table 7.3). Among individual amino acids, percentage increases ranged from 1.8% (methionine) to 7.1% (threonine, cystine, serine), and this response pattern refl ects the relatively higher inherent digestibility of methionine (0.894). There was a signifi cant negative relationship (r = –0.972; P <0.001) in the tabulated mean values between the response (percentage increase) to phytase and the inherent digestibility of amino acids in the control diets. Indeed, the apparently poor response to phytase when chromic oxide has been used may be associated with an overestimation, compared with alternative markers, of the digestibility of amino acids in the control diet (Cowieson and Bedford, 2009). Impact of phytate on protein/amino acid digestibility On the basis of acid-insoluble ash/titanium oxide broiler assays, microbial phytase has a positive infl uence on ileal amino acid digestibility and, axiomatically, phytate has a negative impact. The de novo formation of binary protein–phytate complexes at acidic pH in the stomach of pigs and forestomach of poultry is probably fundamental to the negative impact of phytate. The capacity of phytate to bind protein as both binary and ternary complexes is established and, as described by Rajendran and Prakash (1993), binary complex formation is a biphasic reaction. The polyanionic phytate molecule electrostatically binds with basic arginine, histidine and lysine residues and this initial, rapid step is followed by a slower aggregation of protein and may result in precipitation of the complex. Binary complex formation occurs at a pH less than the isoelectric point of a given protein, and in the case of sodium phytate and α-globulin the reaction was maximal at pH 2.3 and dependent upon phytate to protein ratios. Similarly, sodium phytate interacts with gossypulin, a globulin cottonseed protein, at pH 2.0–3.0 (Yunusova and Moiseeva, 1987). Pivotally, several studies have found that complexed protein is refractory to pepsin hydrolysis (Barré and Nguyen-van-Hout, 1965; Camus and Laporte, 1976; Kanaya et al., 1976; Inagawa et al., 1987; Knuckles et al., 1989). Moreover, Vaintraub and Bulmaga (1991) reported that phytate retarded pepsin hydrolysis of soy protein by 60% at pH 2.0–3.0 under in vitro conditions, but this did not occur at pH 4.0–4.5. These workers concluded that phytate retards pepsin digestion only when phytate is bound to the protein, which makes the important distinction that phytate binds with the substrate (protein) and not the enzyme (pepsin). Indeed, the paucity of basic amino acids in pepsin (Blumenfeld and Perlmann, 1958; Tang et al., 1973) probably precludes phytate from binding with the enzyme. However, although phytate and pepsin may not interact directly, the activation fragment of pepsinogen is heavily basic (13/44 amino acid residues) and so phytate may compromise activation of the zymogen (Dykes and Kay, 1977; Dunn et al., 1978). Alterations in protein solubility and structure induced by aggregation with phytate presumably render the substrate less susceptible to pepsin activity, and thus phytate impedes the initiation of the protein digestive process. Additionally, pepsin-generated peptides are regulators of protein digestion processes (Krehbiel and Matthews, 2003), so it follows that pepsin-refractory complex formation may disrupt these regulatory functions. Although protein–phytate complexes dissociate once gut pH exceeds protein isoelectric points, proteins still may be less readily digested in the small intestine due to structural changes pursuant to their aggregation with phytate. Furthermore, the dissociated complexes release proteins that have escaped pepsin processing and, as a result, are not optimally processed for digestion by trypsin, chymotrypsin, elastase and additional small-intestinal proteases. As a result the rate of protein digestion and absorption is reduced, and if transit rates remain largely unchanged this could result in delivery of excess nitrogen to the fermentative bacteria in the large intestine, with the concomitant risk of multiplication of putrefactive bacteria. Low (1990) concluded that physicochemical properties of foodstuffs are dominant determinants of gastric function and, although speculative, the refractory nature of insoluble protein–phytate complexes may prompt gastric hypersecretion of pepsin and HCl as a compensatory mechanism. Decuypere et al. (1981) investigated the effects of diets containing water-soluble or insoluble soy protein isolates (140 g kg–1) in pigs fi tted with gastric fi stulae. It was concluded that the physical properties of the protein sources were important in regulating pepsin and HCl secretions, as there were marked differences between diets in the 3 h postprandial interval. For example, pepsin secretion with insoluble soy protein was about 88% higher than with soluble soy protein 150 min following feed intake. Zebrowska et al. (1983) reported that pepsin activity in digesta from the proximal duodenum of pigs offered a barley–soybean meal diet was 93% higher than those fed on ‘phytate-free’ diets containing wheat starch, casein and sucrose. The barley–soybean meal diet contained a retrospectively estimated 9.8 g phytate kg–1. Korczynski et al. (1997) offered isonitrogenous, low- (wheat–casein) and high-fi bre (wheat bran– wheat–casein) diets to pigs with denervated gastric pouches. However, the increase in dietary fi bre was associated with an estimated increase in phytate levels from approximately 6.9 to 16.6 g kg–1, and the dietary transition increased pepsin secretion by 70%. Like phytate, condensed tannin also has the capacity to bind protein; therefore, it is relevant that tannin has been shown to increase pepsin and HCl secretion in rats (Mitjavila et al., 1973). It is possible the secretion of the regulatory peptide, gastrin (Burhol, 1982; Furuse, 1999) triggers the compensatory outputs of pepsin and HCl in response to the gastric presence of refractory, phytate-bound protein. As pepsin and HCl are ‘endogenous aggressors’ (Allen and Flemstrom, 2005), their increased outputs would be countered by protective mucin and sodium bicarbonate (NaHCO3) secretions. Importantly, therefore, phytate has been shown to increase excretion of mucin and Na in broilers, which was ameliorated by microbial phytase (Cowieson et al., 2004). As mucin remains largely undigested in the small intestine, any increase in mucin secretion would exacerbate fl ows of endogenous amino acids derived from its protein component. Moreover, it has been demonstrated that phytate increases, and phytase decreases, endogenous amino acid fl ows in broilers (Cowieson and Ravindran, 2007; Cowieson et al., 2008). The amino acid profi les of pepsin (Blumenfeld and Perlmann, 1958; Tang et al., 1973) and mucin (Lien et al., 1997) have been documented and, instructively, the phytase-induced percentage increases in amino acid digestibility in broilers (Table 7.3) are correlated with amino acid profi les of pepsin (r = 0.54; P <0.05) and mucin (r = 0.70; P <0.01). These signifi cant relationships indicate that microbial phytase enhances the digestibility of amino acids that are abundant in pepsin and mucin, presumably via stemming endogenous amino acid fl ows. Ravindran et al. (2006) demonstrated that increasing dietary phytate levels decreased ileal Na digestibility (–0.38 versus –0.24; P <0.05) and, conversely, microbial phytase increased Na digestibility (–0.18 versus –0.52; P <0.001). Also, 500 FTU E. coli phytase kg–1 increased ileal digestibility coeffi cients of Na from –0.52 to –0.04 in broilers offered wheat-based diets containing 11.0 g phytate kg–1 (Selle et al., 2009b). Thus phytate has the capacity to pull Na into the small intestinal lumen, but this depletion of Na is counteracted by phytase. This phytate-induced transition of Na into the gut lumen may be in the form of NaHCO3 to buffer excess HCl secretion. In addition, NaHCO3 has been shown to enhance intestinal alkaline phosphatase activity in rats (Akiba et al., 2007), which may be another reason for the movement of Na into the gut in response to dietary phytate. Furthermore, work by Mothes et al. (1990) demonstrated that the rate of formation of protein–phytate complexes could be reduced signifi cantly through the addition of increasing levels of Na, with levels equivalent to those found in a 0.2% Na diet being suffi cient to break up more than 65% of these complexes. It is possible, therefore, that the current ‘requirements’ for Na, which were generated prior to the use of phytases, may encompass a need to have adequate gastric Na concentrations to displace protein–phytate complex formation. In the presence of increasing phytase dosage, such complex-disrupting activities become more and more superfl uous and, as a result, the Na requirement of the animal may well need to be reviewed in this era of ubiquitous phytase usage. The absence of Na+ in the medium has been shown to inhibit arginine, glutamic acid, glycine, leucine and valine transport in avian intestinal tissue (Lerner, 1984). Also, Ravindran et al. (2008) found that phytase increased ileal amino acid digestibility in maize–soy broiler diets at low dietary Na levels, but that this effect was diminished with increasing NaHCO3 inclusion. It follows that phytate-induced Na depletion in the small intestine may disrupt Na-dependent transport systems and sodium pump (Na+-K+-ATPase) activity and this, in turn, could lead to diminished intestinal uptakes of amino acids and other nutrients. Phytate, as sweet potato extracts or Na phytate, has markedly reduced jejunal and ileal Na+-K+-ATPase activity in rats (Dilworth et al., 2005). Alternatively, 1000 FTU E. coli phytase kg–1 increased Na+-K+-ATPase activity in the duodenum and jejunum of broilers offered maize–soy diets by nearly 20% (Liu et al., 2008). Also, phytate has been shown to reduce intestinal uptakes of glutamic acid and leucine as free amino acids in chickens (Onyango et al., 2008). In summary, the likelihood is that phytate decreases protein digestibility, exacerbates endogenous amino acid fl ows and depresses intestinal uptakes of dietary and endogenous amino acids. The amelioration of these infl uences by phytase may be expressed as increased ileal amino acid digestibility in broilers. Microbial phytase amino acid digestibility assays in swine While there is the possibility that phytase increases ileal amino acid digestibility in pigs, the majority of assays indicate that this is not the case. This is curious because some studies have suggested that phytase enhances protein utilization in pigs (Beers and Jongbloed, 1992; Ketaran et al., 1993; Campbell et al., 1995; Biehl and Baker, 1996; Selle et al., 2003a). The Ketaren et al. (1993) study is of particular relevance because A. niger phytase increased protein deposition by 13.9% (123 versus 108 g day–1) and protein retention by 9.1% (0.36 versus 0.33) in grower pigs. However, these improvements may have been secondary to phytase enhancing skeletal development rather than solely a primary ‘protein effect’.Interestingly, Just et al. (1985) offered a range of diets with an average protein level of 161 g kg–1 to 50 kg female pigs and determined protein deposition rates, which averaged 85 g day–1. From retrospective estimates, 21 diets had average phytate contents of 9.4 g kg–1 and phytate:protein ratios of 0.063:1.0. However, dietary phytate:protein ratios were negatively correlated to protein deposition rates (r = –0.48; P <0.05). The linear regression equation is as follows: Protein deposition (g day–1) = 99.8 – (243 × phytate:protein ratio). The Just et al. (1985) study therefore suggests that increases in phytate levels relative to dietary protein have a deleterious impact on protein deposition. Also, the above equation predicts that a reduction in the phytate:protein ratio, via phytase degrading 50% of dietary phytate, would increase protein deposition by 8.3%, from 85.6 to 92.7 g day–1. Nevertheless, in general microbial phytases have not generated ileal amino acid digestibility responses of corresponding magnitudes. Only three studies (Offi cer and Batterham, 1992a,b; Barnett et al., 1993; Kornegay et al. 1998) have been reported where phytase has tangibly enhanced ileal amino acid digestibility. Coincidentally or not, in these three studies ileal digesta samples were taken from ‘intact’ (slaughtered or anaesthetized) pigs rather than cannulated animals. The method for taking ileal digest samples, either by various cannulation procedures or directly from intact pigs, appears to be pivotal. As reviewed by Selle and Ravindran (2008), in fi ve assays involving 61 observations, phytase increased AID coeffi cients of amino acids by an average of 6.5% (0.767 versus 0.723) at a mean inclusion rate of 590 FTU kg–1 in intact pigs. In contrast, in cannulated pigs, 905 FTU phytase kg–1 increased AID of amino acids by only 1.7% (0.798 versus 0.785) from 281 observations in 11 studies. It is noteworthy that chromic oxide was used as the dietary marker in all 16 studies. Curious outcomes have arisen from phytase amino acid digestibility assays in cannulated pigs. For example, Mroz et al. (1994) reported that 800 FTU A. niger phytase kg–1 signifi cantly increased AID of methionine by 5.1% (0.806 versus 0.767), but numerically depressed threonine digestibility by 2.4% (0.720 versus 0.738). This response pattern is quite unusual as, among essential amino acids, threonine is usually the most, and methionine the least, phytase responsive. A more typical pattern was reported by Kornegay et al. (1998) in intact pigs, where phytase improved threonine digestibility by 16.2%, but methionine digestibility by 9.3% (Table 7.4). In intact pigs, Offi cer and Batterham (1992a,b) reported that microbial phytase substantially increased the ileal digestibility of ten amino acids in linola meal (400 g kg–1) by an average of 14.5% (0.715 versus 0.627), as shown in Table 7.4. Linola meal, a variant of linseed meal, was the sole protein source and the grower pigs were fed on a once-daily basis. Increasing Ca concentrations, relative to dietary phytate and protein levels, may depress amino acid digestibility responses to phytase (Selle et al., 2009a). In the studies of Offi cer and Batterham (1992a,b), the basal diet contained approximately 8.8 g Ca kg–1 and 9.0 g total P kg–1; thus the ‘inverse’ Ca:P ratio of 0.98 coupled with Factors infl uencing protein–phytate complex formation Factors that infl uence protein–phytate complex formation are considered because the extent of their formation is probably critical to the ‘protein effect’ of phytate and phytase. The results from the in vitro study of Vaintraub and Bulmaga (1991) emphasize the critical importance of pH on the pepsinrefractory nature of complexed protein. At pH 2.5, pepsin digestion of phytatebound casein was retarded by 50% but the digestion of casein was not impeded at pH 4.0. Gizzard fl uid taken from 22 non-anaesthetized birds had an average pH of 2.05 (Farner, 1943), which would be conducive to phytate binding protein and reducing its vulnerability to pepsin digestion. Phytate has an affi nity for casein, as Na phytate has been shown to reduce in vitro casein solubility from 99 to 1% at pH 2.0 (Kies et al., 2006). Shan and Davis (1994) added 20 g Na phytate kg–1 to an atypical broiler diet containing 150 g casein kg–1, which depressed weight gain (44%), feed intake (22%) and feed effi ciency (29%) from 28 to 46 days post-hatch. Presumably the profoundly depressed growth performance was pursuant to reduced protein digestibility following the binding of casein by Na phytate. Given the importance of pH in the stomach or proventriculus, it is noteworthy that limestone, a key source of Ca, has a very high acid-binding capacity of capacity of 15,044 meq kg–1 at pH 3.0 (Lawlor et al., 2005). Thus Ca, as limestone, will tend to increase gut pH and high dietary limestone levels may depress the formation of protein–phytate complexes. The propensity of proteins to be bound by phytate is variable, which may be dependent on their structure and the accessibility of basic amino acid residues (Champagne, 1988). For example, Kies et al. (2006) found that the affi nity of phytate for canola meal protein was relatively low. At pH 2.0, Na phytate reduced the solubility of canola meal protein solubility from 100 to 63% but phytate had little infl uence as pH increased. This is consistent with the relatively modest average increase of 2.7% (0.799 versus 0.778) in AID coeffi cients of 14 amino acids following the addition of A. niger phytase 1200 FTU kg–1 to a broiler diet containing 526 g canola meal kg–1 (Ravindran et al., 1999a). Alternatively, Kies et al. (2006) reported that Na phytate reduced soybean meal protein solubility from 91 to 2% at pH 2.0 and from 60 to 23% at pH 3.0. In keeping, Ravindran et al. (1999a) reported that phytase increased amino acid AID coeffi cients by a more robust 4.2% (0.850 versus 0.816) in broiler diets containing 438 g soybean meal kg–1.Ravindran et al. (1999a) also reported that phytase enhanced AID of amino acids in wheat (9.3%) to a greater extent than in maize (3.4%) in broilers, and this difference in response to phytase was subsequently confi rmed (Rutherfurd et al., 2002). These fi ndings are consistent with the formation of protein–phytate complexes in wheat reported by Hill and Tyler (1954b), whereas O’Dell and De Boland (1976) did not detect protein–phytase complex formation in maize. However, Kies et al. (2006) reported that Na phytate reduced the solubility of maize protein from 100 to 28% at pH 2.0, with a more modest reduction at pH 3.0 and little infl uence at pH 4.0–5.0. However, the negative fi nding by O’Dell and De Boland (1976) was made following gel fi ltration at pH 4.4, which does not preclude phytate complexing maize proteins at a more acidic pH. It seems reasonable to conclude that, if the pH that prevailed in the forestomach of broilers offered maize-based diets in the studies of Ravindran et al. (1999a) and Rutherfurd et al. (2002) had been more acidic, both complex formation and amino acid digestibility responses to added phytase might have been greater. Broilers may be fed diets containing a proportion of whole grains, which stimulates gizzard function (Cumming, 1994). In one experiment, it was shown that feeding whole grains signifi cantly increased gizzard weight by 28% (37.5 versus 29.2 g) and reduced the pH of gizzard digesta from 3.6 to 2.9 (Rutkowski and Wiaz, 2001). On the one hand, this reduction in pH would tend to increase the solubility of Mg–phytate complexes (Cheryan et al., 1983) and presumably increase the extent of phytate degradation by exogenous phytase, particularly if digesta are retained for longer intervals in a more active gizzard. On the other, the formation of insoluble protein–phytase complexes would be favoured by such a pH reduction. Therefore, microbial phytase may be more effective in a context of feeding whole grains as opposed to diets in which the entire grain component is ground. ‘Energy Effect’ of Phytase The possibility that phytate depresses energy digestion and utilization and that phytase has a reciprocal, positive effect is clearly an increasingly important issue. Microbial phytase consistently enhances metabolizable energy (ME) of broiler diets, but the impact of phytase on digestible energy (DE) of pig diets is not as pronounced. One example where a DE effect has been reported in pigs (from 20–107 kg liveweight) is that of Johnston et al. (2009). The effect of phytase on digestibility and subsequent utilization of energy suggests that net energy (NE) studies may provide clarifi cation. The formulation of pig diets on the basis of NE (de Lange and Birkett, 2005) is an increasing practice. It seems possible that the phytate content of relevant feedstuffs contributes to the differential between the NE of a diet and the DE of swine diets and the ME of poultry diets. Certainly, the work of Pirgozliev et al. (in press) suggests that, in poultry, in some cases the use of phytase has little effect on ME but signifi cant effects on NE, suggesting there may be a post-absorptive partitioning effect of this enzyme. On the other hand, the consistent benefi cial effect that phytase daily energy intake. The use of phytase under these circumstances will increase the proportion of AME intake that is in excess of maintenance requirements. Thus NE studies, or AME studies coupled with data on intake effects of this enzyme, would be most appropriate. The vast majority of studies, however, have focused on the DE and AME effects of this enzyme in isolation. Early studies in poultry, involving dephytinized feed ingredients, suggested that phytate negatively infl uences energy utilization in broilers (Rojas and Scott, 1969; Miles and Nelson, 1974). As reviewed by Selle and Ravindran (2007) in a series of 12 studies, phytase activity of 662 FTU kg–1 increased the AME of broiler diets by an average of 0.36 MJ (13.64 versus 13.27 MJ kg–1 dry matter) where the percentage responses in AME were negatively correlated (r = −0.562; P <0.02) to the energy density of the control diets.