**2.2. Branched-chain amino acids**

Knowledge of the branched-chain amino acids concerning poultry started in much the same way as arginine with the determination of essentiality for poultry in 1944 [12, 13], and the first data set outlining their requirements was published two years later in 1946 [41]. The original requirement values were established to be 0.5, 1.5, and 0.7% of the diet for isoleucine, leucine, and valine, respectively, and varied little during the years of experimentation contributing to crystalline amino acid diets [14, 16, 19, 42].

In 1960, Laksesvela [43] reported that that deletions of isoleucine resulted in a 27% reduction in the "combinative protein value" of herring solubles, whereas additions of leucine resulted in a 16% reduction in the aforementioned metric. The implications of this discovery would not be fully appreciated until 1968 when Mathieu and Scott [44] reported that feeding excess leucine in diets containing isoleucine and valine near adequacy resulted in depressions in body weight. This report started investigations into branched-chain amino acid antagonism, as the existence of amino acid antagonisms were known (i.e., the existing work on the lysine-arginine antagonism) as well as its previous discovery in rats [45].

The second interaction investigated in the early 1970s by D'Mello and Lewis in a series on amino acid interactions in chick was that among the branched-chain amino acids [46]. As with lysine and arginine, D'Mello and Lewis sought to confirm the existence of the former amino acid antagonism in chick nutrition. Through the course of five experiments, D'Mello and Lewis [46] isolated and definitively showed the existence of an antagonism between leucine and isoleucine and leucine and valine, but theorized that up to six antagonisms existed among the branched-chain amino acids. D'Mello and Lewis [46] further defined that this antagonism is most prevalent when valine and isoleucine are included in the diets at adequate levels but could present itself if valine was not the limiting amino acid in the basal diet, in the case of a leucine-valine antagonism.

In their next series of experiments, D'Mello and Lewis [34] determined the influence of excess leucine on the requirements of valine and isoleucine. Excess leucine shifted the requirements of both valine and isoleucine in order to obtain maximal body weight gain (**Figure 3**). Of particular interest, D'Mello and Lewis [46] had indicated that the leucine × valine interaction was probably of more practical importance to broiler production and later showed that adjusting valine to maximize growth at higher leucine levels improved average daily gain above control levels whereas

**Figure 3.**

*Influence of titrating isoleucine and valine on average daily gain at low (solid line), medium (dashed line), and high leucine (dotted line). Adapted from D'Mello and Lewis [34].*

increasing isoleucine resulted in a decrease in optimal average daily gain [34]. In their final study, D'Mello and Lewis [47] established the existence of a metabolic mechanism behind the branched-chain amino acid antagonism by pair feeding diets containing adequate and excessive leucine, as depressions in body weight gain were still observed in birds fed diets containing excess leucine compared with consuming an equal amount of a control diet.

The following year, Allen and Baker [48] conducted a series of experiments to determine the efficacy of isoleucine and valine when leucine was fed in excess. The ability of isoleucine to sustain body weight gain was linearly reduced to 80% of control levels when leucine was increased from 0 to 3% of the diets, conversely the ability of valine to support body weight gain was quadratically reduced to 74% of control levels as leucine supplementation was increased to 6%. Due to the difference in leucine excesses employed in the isoleucine and valine experiments, the minimal efficacy values do not allow for a direct comparison. Equalizing leucine inclusion levels to 3% displays an average efficacy of 81 and 79 for isoleucine and valine, respectively, agreeing with the postulations of D'Mello and Lewis [46] that the valine × leucine interaction would likely have more impact on poultry production.

Due to the similarities in chemical structure and common enzymes used in the transamination and decarboxylation steps of catabolism [49, 50], researchers believed that the antagonism among the branched-chain amino acids was linked to increased catabolism brought about by excessive leucine. Researchers at the University of Nottingham tested this hypothesis by monitoring the activity of aminotransferase for leucine and valine [51], as well as the catabolism of C14 labeled valine [52]. Both studies failed to observe any influence on the rate of catabolism of valine when leucine were fed in excess. Conversely, Smith and Austic [53] observed a small increase, approximately 2% of ingested levels, in the catabolism of C14 labeled valine and isoleucine when leucine reached 2.25% of the diet. Similarly, Calvert et al. [54] observed a 50 and 43% increase in isoleucine and valine, respectively, when leucine was fed at 5%. In addition, Calvert et al. [54] pair fed chicks diets containing 1.2 or 5.0% leucine to gauge the effect of reduced feed intake on branched-chain amino acid antagonism responses. Calvert et al. [54] found that growth depressions associated with excess leucine persisted when feed intake was equalized, agreeing with the previous findings of D'Mello and Lewis [47]. Based on their overall findings, Calvert et al. [54] proposed that 70% of the negative effects associated with branched-chain amino acid antagonism is linked to feed intake as opposed to a primary effect of metabolic changes.

Jackson and Potter [55] reported that the classic responses of branched-chain amino acid antagonism observed in poultry also occurred in turkeys. Branched-chain amino acid antagonism had previously observed in turkey poults [56, 57], but Jackson and Potter [55] also discovered that a reciprocal antagonism between isoleucine and valine that could result in depressions in body weight when either was fed at adequacy while the other was fed in excess. Mendonca and Jensen [58] latter confirmed the existence of isoleucine and valine antagonism in chickens, when it was found that supplementing isoleucine reduced performance, whereas a concomitant addition of isoleucine and valine had no effect.

Unlike with lysine-arginine antagonism, less is know about the mode of action behind the branched-chain amino acids with many theories lacking critical evidence to definitively prove. Despite this, cornerstone data were generated during the research conducted from the 1960s to 1980s. Firstly, the branched chain amino acid antagonism is a reciprocal antagonism, in that it can present itself by targeting both valine and isoleucine, and subsequent interactions between valine and isoleucine. Secondly, leucine appears to be the primary antagonist, but apparent performance gains can be made if proper supplementation of valine is made to account for the antagonism. Lastly, the largest piece of information that can be gleaned from classic research is that the antagonism is most apparent when isoleucine and valine are at adequacy levels, indicating that negative effects are likely to occur in reduced crude protein diets.
