In standard broiler diets, protein-bound amino acids are usually derived from soybean, Canola and other protein meals and cereal grains including maize, wheat and, to lesser extents, barley and sorghum. The amino acid digestibilities of these feedstuffs are well established, which may be impacted by a range of anti-nutritive factors including phytate and soluble NSP. The effects of both can be counteracted by exogenous enzymes and, indeed, exogenous phytase can have profound effects on the digestive dynamics of protein-bound amino acids. Phytase (500 FTU/kg) has been shown to increase average apparent digestibility coefficients of 16 amino acids by 49.7% (0.720 versus 0.481) in the proximal jejunum in broilers offered maize-based diets [60]. The implication is that phytase triggered a ‘proximal shift’ in the site of intestinal uptakes of essentially protein-bound amino acids. Moreover, there are indications that rapidly absorbed protein-bound amino acids will advantage broiler growth performance and influence the post-enteral availability of amino acids [61]. An initial assessment of amino acid digestive dynamics derived from protein-rich feedstuffs for poultry has been completed [62] and the generation of more, similar data should be a priority.

Oligopeptides are defined as di- and tri-peptides; importantly, the majority of amino acids are absorbed as oligopeptides as opposed to monomeric amino acids [63]. Moreover, the likelihood is that intestinal uptakes of oligopeptides, principally via the PepT-1 transporter [64], are more rapid and efficient than monomeric amino acids [65, 66]. An early demonstration found that when glycine was administered as a dipeptide in humans, 30 minute post-prandial glycine plasma levels were 94.7% higher (4.40 versus 2.26 mg/100 mL) than monomeric glycine [67]. More recently, it was reported that peptides derived from soy 11S globulin were more rapidly and efficiently absorbed in adult males than soy 11S globulin protein or a corresponding mixture of monomeric amino acids [68].

The dipeptide, carnosine ($\beta$-alanyl-L-histidine) and its methylated analogue, anserine, have been exhaustively reviewed [69]. Carnosine a naturally occurring dipeptide of special interest both as an entity and by virtue of its histidine and alanine components [70] and high carnosine concentrations are found in breast muscle [71]. Dietary inclusions of histidine (1.8 g/kg), alanine (3.2 g/kg), individually and in combination, plus carnosine (2.7 g/kg) in diets based on a maize-wheat blend were compared in [72]. At 28 days post-hatch, carnosine generated significantly heavier absolute breast weights than the negative control diet by 19.5% (207.7 versus 173.8 g/bird) and higher breast yields by 1.6 percentage units (16.2 versus 14.6%) and carnosine was numerically superior to the balance of treatments based on its constituent monomeric amino acids. Interestingly, reduced carnosine concentrations in breast muscle have been associated with myopathies such as wooden breast and white striping [73], which is an issue of increasing concern. Lackner et al. [74] claimed that additional histidine has the potential to reduce the incidence of white striping, but carnosine may be a better way to fortify reduced-CP diets with histidine and alanine than their provision as non-bound amino acids. Given the above, it follows that carnosine merits further investigation.

Non-bound (synthetic, crystalline, feed-grade) amino acids are alternative ‘protein’ sources to soybean meal and other relevant feedstuffs in broiler diets [20]. However, an important distinction in that intestinal uptakes of non-bound amino acids are more rapid than protein-bound amino acids because prior protein digestion in the gut lumen is not required [75]. This was demonstrated decades ago; specifically, with methionine in rats [76] and subsequently with lysine and threonine in pigs [21].

The digestion rates of protein-bound of 13 amino acids in conventional sorghum-based diets were determined by Liu et al. [77], as shown in Fig. 1. The mean digestion rate was 2.35 $\times$ 10${}^{-2}$ min${}^{-1}$, which ranged from 1.85 (tyrosine) to 3.25 $\times$ 10${}^{-2}$ min${}^{-1}$ (arginine). Collectively, the branched-chain amino acids’ digestion rate averaged 2.05 $\times$ 10${}^{-2}$ min${}^{-1}$; this slow rate reflects their hydrophobicity. Methionine, lysine and threonine were present as both non-bound and protein-bound amino acids and their digestion rates are compared with exclusively protein-bound amino acids in Fig. 2. The sorghum-based diets contained 4.8 g/kg lysine HCl and 3.4 g/kg dl-methionine; thus, non-bound lysine and methionine represented 27.9% and 51.5% of their specified dietary levels. The overall lysine and methionine digestion rates were 4.14 $\times$ 10${}^{-2}$ min${}^{-1}$ and 5.51 $\times$ 10${}^{-2}$ min${}^{-1}$, respectively. Then, it may be deduced that the digestion rate of non-bound lysine was 8.78 $\times$ 10${}^{-2}$ min${}^{-1}$ and non-bound methionine 8.49 $\times$ 10${}^{-2}$ min${}^{-1}$, based on the assumption that the rate of the protein-bound component was 2.35 $\times$ 10${}^{-2}$ min${}^{-1}$. Thus, non-bound amino acids have more rapid intestinal uptakes than protein-bound amino acids by nearly a four-fold factor.

Fig. 1.

The digestion rate constants (DRC $\mathrm{\times }$ 10${}^{\mathrm{-}\mathrm{2}}$ min${}^{\mathrm{-}\mathrm{1}}$) of 13 protein-bound amino acids in broiler chickens offered sorghum-based diets; adapted from Liu et al. [77].

Fig. 2.

The digestion rate constants (DRC $\mathrm{\times }$ 10${}^{\mathrm{-}\mathrm{2}}$ min${}^{\mathrm{-}\mathrm{1}}$) of methionine, lysine, threonine and the mean of 13 protein-bound amino acids in broiler chickens offered sorghum-based diets; adapted from Liu et al. [77].

Given these disparities, it is worth exploring the possibility of enteric coating one or more monomeric amino acids so that their intestinal uptake rates are retarded. Advanced oral delivery technologies facilitating the protection and absorption of protein and peptide molecules are available [78]. Moreover, it has been reported that both encapsulated lysine and methionine were more effectively utilised in poultry than as standard, non-bound entities [79].

The intestinal uptakes of monomeric amino acids take place via an array of Na${}^{+}$-dependent and Na${}^{+}$-independent transport systems with over-lapping affinities for amino acids [80, 81]. Consideration has been given to the likelihood that there is competition between certain amino acids for intestinal uptakes via relevant transporters [82, 83, 84]. More recently, Chrystal et al. [19] reported a quadratic relationship (r = 0.755; p 0.001) between methionine and phenylalanine digestibility coefficients which suggested that methionine was competitively inhibiting phenylalanine intestinal uptakes [85]. One Na${}^{+}$-dependent brush border transporter, PHE, has specific affinities for just these two amino acids [80], which indicates that this was the focus of competition between methionine and phenylalanine for intestinal uptakes.

Intestinal uptakes of glucose and amino acids via Na${}^{+}$-dependent transporters involve co-absorption with sodium (Na). Glucose and Na are principally absorbed via SGLT-1 [86], whereas, amino acids and Na are absorbed by a range of about fourteen Na${}^{+}$-dependent transporters [79]. The likelihood of competition between glucose and amino acids for intestinal uptakes has been discussed by several researchers [87, 88, 89, 90]. However, Moss et al. [33] found negative Pearson correlations between apparent digestibility coefficients of starch and 12 of 16 amino acids assessed in the proximal ileum. The significant correlations ranged from serine (r = –0.327; p = 0.023) to phenylalanine (r = –0.507; p 0.001). Additionally, negative correlations were also observed in the proximal jejunum, distal jejunum and distal ileum. It should be noted that the maize-based experimental diets in [33] contained unusually high starch concentrations which would have contributed to the declaration of these negative Pearson correlations. Nevertheless, the relatively high starch concentrations in reduced-CP diets may be negatively impacting on amino acid digestibilities to some extent.

The addition of lysine HCl to increase the dietary lysine concentrations from 10.0 to 11.8 g/kg in broiler diets has been shown to increase significantly the apparent ileal digestibility of lysine per se but also the digestibility of seven additional amino acids: isoleucine, methionine, phenylalanine, valine, aspartic acid, glutamic acid, tyrosine [91]. The average digestibility of the eight amino acids involved increased by 3.91% (0.824 versus 0.793) and of lysine per se by 5.42% (0.837 versus 0.794). This was attributed to dietary lysine enrichment up-regulating lysine transport systems with over-lapping amino acid specificities along the small intestine, based on data generated by Torras-Llort et al. [92]. That the moderate dietary addition of one non-bound amino acid generated significant differences in the digestibilities of seven other amino acids raises the question as to the magnitude of the impact generated by tangible additions of a range of non-bound amino acids in reduced-CP diets.

There is also evidence that competition for intestinal uptakes amongst di- and tri-peptides via the oligopeptide transporter, PepT-1, take place. A glycine-leucine dipeptide has been shown to depress absorption of a glycine-glycine dipeptide via partially or totally competitive inhibition in human subjects [93]. This evidence for competition between monomeric amino acids and oligopeptides for intestinal uptakes only complicates determinations of apparent amino acid digestibility coefficients as discussed in Section 3.

The visceral tissues, including the intestinal mucosa, pancreas and spleen, have high metabolic rates and the contribution of portal drained viscera to whole body energy expenditure is substantial and dietary amino acids are a major source of mucosal energy generation in pigs [94]. The majority of absorbed glutamic acid (95%), aspartic acid (95%) and glutamine (70%) are catabolised in the gut mucosa and do not enter the portal circulation in pigs, and probably other mammalian species [95]. This also applies to proline (40%) and the branched-chain amino acids (35–40%) to lesser extents. However, very little is known about the catabolism of amino acids or alternative energy substrates by the gut mucosa in broiler chickens, which has a considerable bearing on their bioequivalency. Watford et al. [96] concluded that glucose, glutamine and glutamate are the preferred energy substrates in avian enterocytes and this outcome was confirmed by Porteous [97], who found that glucose, glutamine and glutamate were the only individual substrates to stimulate respiration in intestinal brush-border cells of broiler chickens. There is the implication that a ‘catabolic ratio’ exists in avian enterocytes between glucose and amino acids (predominantly glutamate/glutamine) as substrates for energy provision but the extent that the balance of amino acids are catabolised is not known. In rats, glucose and glutamine provided similar proportions of energy to enterocytes in the gut mucosa [98]. This is consistent with the ‘catabolic ratio’ principle, although there are much higher rates of respiration in the gut mucosa or chickens than rats [97]. In respect of additional amino acids undergoing catabolism, it is pertinent that Stoll et al. [99] reported that catabolism dominated the first-pass intestinal metabolism of essential amino acids in young pigs and that catabolism of amino acids within enterocytes was quantitatively greater than amino acid anabolism or incorporation into mucosal protein. Catabolism of amino acids could result from decarboxylation or deamination and in the second reaction the amino group is removed and converted to ammonia (NH${}_3$) and the net portal outflow of ammonia accounted for 18% of total amino acid nitrogen intake [99].

The pigs were fitted with arterial, venous, portal and gastric catheters plus an ultrasonic portal blood flow probe to determine the net portal NH${}_3$ outflow in the Stoll et al. [99] study. Such surgical interventions may not be feasible for determinations of the net portal flux of amino acids in broiler chickens. However, effectively the gross portal flux of amino acids has been determined in poultry by taking blood samples from the anterior mesenteric vein in three studies. In the first study, a ‘slow’ protein digestion rate foundation diet was compared with a ‘rapid’ protein digestion rate summit diet in broiler chickens from 7 to 28 days post-hatch [57]. The summit and foundation diets had analysed CP concentrations of 207.6 and 207.0 g/kg respectively and the essential differences were increases in non-bound amino acid inclusions (arginine, isoleucine, lysine, methionine, threonine, tryptophan) from 2.77 to 11.25 g/kg collectively, and in casein from 0 to 50 g/kg. Free amino acid concentrations were determined in both the portal (anterior mesenteric vein) and systemic (brachial vein) circulations. Total free amino acid concentrations in the portal circulation exceeded those in the systemic circulation by 14.7% (950 versus 828 $\mu$g/mL) but the patterns of amino acid concentrations were highly correlated (r = 0.998; p 0.0001), which reflects the fact that amino acids in the portal circulation gain entry to enterocytes from both the gut lumen and the arterial circulation to the small intestine [100]. Instructively, free concentrations of six amino acids that were present in diets in part as non-bound amino acids collectively increased by 17.8% (245 versus 208 $\mu$g/mL); in comparison, free concentrations of fourteen protein-bound amino acids decreased by 1.94% (656 versus 669 $\mu$g/mL).

A similar outcome was reported in the second study [33] in which free amino acid concentrations in the portal circulation were determined in birds were offered maize-based diets with specified CP contents of 213 and 178 g/kg from 7 to 28 days post-hatch. The reduction in dietary CP resulted in an increase in non-bound amino acids from 3.54 g/kg (lysine, methionine, threonine) to 10.90 g/kg (lysine, methionine, threonine, isoleucine, valine, arginine). Across these amino acids, collective portal concentrations increased by 33.7% (333 versus 249 mg/mL) following the dietary CP reduction. However, the collective increase for the balance of ten protein-bound amino acids was a more modest 5.28% (738 versus 701 mg/mL).

In the third study [36], free amino acid concentrations in portal plasma in birds offered 215 and 165 g/kg CP, wheat-based diets were determined. Increases in portal concentration following the dietary CP reduction were confined to lysine with an increase of 33.7% (51.6 versus 38.6 $\mu$g/mL), methionine with an increase of 25.2% (14.4 versus 11.5 $\mu$g/mL) and an increase of 25.0% (70.0 versus 56.0 $\mu$g/mL) for threonine. These were the only three non-bound amino acids to be included in both 215 and 165 g/kg CP diets. However, the dietary CP reduction depressed collective plasma concentrations of nine protein-bound only amino acids by 21.1% (160.2 versus 203.0 $\mu$g/mL).

The consistent contrasts in these three studies suggest that non-bound amino acids are more likely to gain access to the portal circulation than their protein-bound counterparts. That non-bound amino acids enter anabolic and/or catabolic pathways within enterocytes to lesser extents may be due to their more rapid intestinal uptake rates and proximal sites of absorption in the small intestine where more starch/glucose is available as an alternative energy substrate.

The entries of amino acids into the portal and, via the liver, systemic circulations are fragmented by their digestibilities, rates of intestinal uptakes and transitions across the gut mucosa which promotes the likelihood of post-enteral amino acid imbalances [101]. Consideration was given to the short-term dynamics in amino acid metabolism in this context by Schreurs et al. [102], who proposed that circulating free amino acids could be either incorporated into protein or undergo metabolic degradation. “Postprandial oxidative losses” was the description given to the second fate and was quite extensively evaluated by Victor Schreurs and his colleagues. The [${}^{13}$CO${}_2$] breath test was used to investigate short-term postprandial amino acid catabolism in rats [103], in which egg white protein and a corresponding blend of non-bound amino acids were compared. The postprandial catabolism of radioactively labelled leucine in the non-bound amino acid diet was more rapid and of a greater magnitude than in the protein-bound egg white diet, which was attributed to the faster, more proximal absorption of non-bound leucine. Subsequently, Nolles et al. [104] compared postprandial oxidative losses of non-bound and protein-bound amino acids in rats and humans, again using [${}^{13}$CO${}_2$] breath tests. Postprandial oxidative losses were significantly higher for non-bound leucine than protein-bound leucine in egg white in rats after a five-day adaptation period. A similar finding was reported in human subjects. These outcomes are supported by an earlier study [105], which found that protein-bound amino acids in casein supported better whole-body protein homeostasis in rats than a corresponding blend of non-bound amino acids. These researchers suggested that the molecular form of dietary amino acids influences both their post-enteral oxidation and availability for protein synthesis.

Catabolism of amino acids can involve deamination with the release of ammonia (NH${}_3$), which is inherently toxic [106]. Moreover, there are indications that broiler chickens are more susceptible to NH${}_3$ intoxication than mammalian species. The intravenous LD${}_{50}$ of ammonium acetate in poultry is half that of mice (2.72 versus 5.64 mmol/kg) as reported in [107]. Postprandial oxidation of amino acids invites the ‘costs of deamination’ [108] and energy costs of deamination may amount to 9.5% of the dietary energy intake from data generated in [109], which is a substantial impost. In addition, the hepatic oxidative deamination of surplus amino acids generates NH${}_3$ that demands detoxification [106]. Ammonia is detoxified by its condensation with glutamic acid to generate glutamine in a reaction catalysed by glutamine synthetase [110]. Ultimately NH${}_3$–N is excreted as uric acid-N in urine following the entry of glutamine into the Krebs uric acid cycle in which, pivotally, one mole of glycine is required for every mole of uric acid excreted [111]. Importantly, Ngo et al. [112] concluded that dietary glycine deficiencies limit uric acid synthesis and the importance of glycine in reduced-CP diets is well established [113]. Indeed, the glycine requirements for uric acid synthesis has been estimated to average of 49.2% of dietary glycine intake, but this requirement was highly variable [114].

Several poultry studies [115, 116, 117] have associated increased NH${}_3$ plasma concentrations with compromised growth performance. The transition from 220 to 190 g/kg CP diets in Ospina-Rojas et al. [117] increased plasma NH${}_3$ concentrations by 59.4% (7.27 versus 4.56 mg/dL), which was associated with a 14.1% (781 versus 909 g/bird) reduction in weight gain and FCR were compromised by 9.79% (1.57 versus 1.43) from 1 to 21 days post-hatch. Thus, it appears that high dietary non-bound amino acid inclusions will generate post-enteral amino acid imbalances because of their relatively rapid intestinal uptakes resulting in the deamination of surplus amino acids. Thus, high non-bound amino acid inclusions in reduced-CP diets may be effectively squandered to some extent and cannot completely meet amino acid requirement targets. Moreover, excessive deamination of imbalanced amino acids coupled with inadequate NH${}_3$ detoxification may result in NH${}_3$ toxicity, or NH${}_3$ overload, to the detriment of bird performance. In support of this proposal, Namroud et al. [115] concluded that large inclusions of non-bound essential amino acids in reduced-CP diets increased blood and excretory NH${}_3$ ammonia concentrations, which may be the main cause of retarded growth and appetite in diets with less than 190 g/kg CP, due to negative effects of NH${}_3$ on tissue metabolism.

Protein accretion, or net protein synthesis, is the difference between protein deposition and protein degradation mainly in skeletal muscles and this protein turnover is a dynamic, continuous process [118]. The rapid accumulation of breast-muscle protein in rapidly growing chicks is due to a marked decrease in the fractional rate of degradation rather than increases in deposition fractional rates [119]. The dynamic nature of protein turnover in skeletal muscle is illustrated in the Urdaneta-Rincon and Leeson [120] study. These researchers examined protein turnover in Pectoralis major in 21-day-old broiler chickens offered fourteen diets with graded levels of lysine and crude protein. The absolute synthesis and breakdown rates of protein were determined and the differences represents net protein synthesis or accretion. Absolute protein synthesis averaged 626 mg/day, protein breakdown averaged 321 mg/day and net protein synthesis averaged 304 mg/day. Moreover, increasing net protein synthesis rates appeared to be quadratically associated with improvements in FCR.

Protein synthesis demands energy as an input of 5.35 kJ is required for the synthesis of one gram of protein [121] and is more energy demanding than protein degradation [122]. As concluded by Moughan [56], protein synthesis efficiency is markedly influenced by the harmonious provision of amino acids and non-amino energy supplying compounds (essentially glucose) to sites of protein synthesis. This emphasises the importance of starch-protein digestive dynamics [123]; however, this topic is peripheral to the remit of this review. The amino acid composition of breast and thigh meat, the two critical portions of the carcass, have been determined [124] and, clearly, the full complement of essential and non-essential amino acids must be made available at sites of protein synthesis to maximise protein accretion which underscores the importance of amino acid bioequivalency. have been studied by 14C02 breath tests, according

As discussed, protein-bound, branched-chain amino acids (BCAA), isoleucine, leucine and valine, are slowly absorbed along the avian small intestine because of their hydrophobicity [77]. As a result, their dietary inclusions as non-bound amino acids generate larger than normal differentials in intestinal uptake rates. In a recent, as yet unpublished, study [125], elevated BCAA levels were assessed in 187.5 g/kg CP broiler diets based on wheat, sorghum or a wheat-sorghum blend from 7 to 28 days post-hatch. A feed grain by BCAA treatment interaction (p 0.001) for weight gain was observed; elevated BCAA levels increased weight gain by 9.26% (1451 versus 1328 g/bird) in birds offered sorghum-based diets but decreased weight gain by 9.49% (1288 versus 1423 g/bird) in their wheat-based counterparts. This is a remarkable outcome and elevated BCAA levels did not influence weight gain in birds offered blended diets. Wheat-based diets contained up to 52.6% non-bound BCAA as a proportion of total analysed BCAA dietary concentrations, whereas the corresponding proportion of 27.2% in sorghum-based diets was substantially less. These proportions were quadratically related to weight gain (r = 0.700; p 0.0001) across all treatments and the relevant quadratic equation predicts that weight gains deteriorate from a maximum of 1428 g/bird once the proportion of 29.2% is exceeded.

In [125], leucine inclusions were elevated from 105 to 150 relative to lysine (100) without and with elevations in isoleucine (65 to 75) and valine (75 to 85) to meet the possibility of BCAA antagonisms. In diets with all three elevations, sorghum-based diets contained 4.93 g/kg non-bound leucine with a balance of 35.81 g/kg non-bound amino acids and wheat-based diets contained 9.12 g/kg non-bound leucine and 45.04 g/kg non-bound amino acids, both in 187.5 g/kg CP diets. Clearly, there are differences in these respects; however, if non-bound and protein-bound amino acids were bioequivalent the remarkable outcome of a 9.26% increase in weight gain in birds offered sorghum-based diets as opposed to a 9.49% decrease with wheat-based diets would not have been likely.

Leucine is usually included in broiler diets at 109 relative to lysine and positive responses to elevated leucine levels in maize-based, broiler diets have been reported [126, 127]. When leucine plasma and intracellular levels exceed requirements for protein accretion, leucine can stimulate protein synthesis by activating the mechanistic target of rapamycin (mTOR) to promote protein deposition and suppress protein degradation [128, 129]. It appears that this may have taken place in sorghum-based diets but not in wheat-based diets. In addition to inherent differences between the two feed grains, one possible contributing factor may have been higher levels of post-prandial oxidation of non-bound leucine (and other amino acids) in birds offered wheat-based diets. The kinetics of radioactively labelled leucine in human subjects was investigated by Metges et al. [130]. These researchers found that resulting from oxidation of non-bound leucine, protein-bound leucine (in casein) was more efficiently used for protein synthesis during the absorptive phase than a casein-equivalent intake of non-bound amino acids. They also concluded that the form in which leucine is consumed affects its immediate metabolic fate and retention in the body. Thus, the outcomes of the Greenhalgh et al. [125] study support the concept that non-bound and protein-bound amino acids are not bioequivalent, as forecast in [131] and this may particularly apply to branched-chain amino acids.