Metabolic Modifiers: Effects on the Nutrient Requirements of Food-Producing Animals (1994)

As discussed for other species, the basic strategies for influencing growth and development of avian species by modifying metabolism to decrease body fat deposition and increase protein accretion include (1) exogenous administration of species-specific somatotropin (ST), in this case chicken ST (cST); (2) exogenous administration of releasing factors controlling endogenous ST secretion; (3) treating the target species with β-agonists administered orally; and (4) various immunological strategies. In addition, avian species demonstrate the phenomenon of true compensatory growth, which in itself influences growth and development patterns.

With limited literature available for citation regarding application of these strategies in avian species, explicit nutrient requirement recommendations are, at best, tenuous. Rather, the approach selected was to briefly review the strategies from a practical perspective as they influence avian growth, citing selected literature, and to suggest a standardized diet that would be appropriate for a range of treatment strategies intended to modify metabolism and body composition. Stated alternatively, the standardized formulation would assure that nutrient intake does not limit biological responsiveness.

Practicality aside, academic interest in the use of pituitary and recombinantly derived cST (see Figure 2-1) has been justified as a potential growth promotant and partitioning tool to overcome the increasing obesity problem in broilers. For example, the proportion of broiler production costs expended in deposition of the abdominal fat pad, a by-product of no consumer value, accounts for fiscal cost of $94 million annually for the 20 leading poultry integrators in the United States.

Exogenous administration of ST from mammalian sources is ineffective as a growth promotant in aves (Libby et al., 1955; Glick, 1960; Sell and Balloun, 1960) with the possible exception of a tryptic digest of bovine ST (bST) (Myers and Petterson, 1974). Questionable purity of such preparations, in addition to a hierarchy of ST actions across species, confound interpretation. Buonomo and Baile (1988) have found that high doses of recombinantly derived bST increase circulating levels of insulin-like growth factor-I (IGF-I) and give a transient increase in growth rate and feed consumption with no change in feed efficiency. They found that attenuation of these responses was associated with high levels of circulating antibodies against the heterologous hormone.

Using pituitary cST characterized chemically and biologically, Leung (1986) reported the response of pituitary-intact broilers. At doses of 5, 10, or 50 µg cST/day given intravenously over a 14-day period in 28-day-old birds, a transitory stimulation (20.6 percent diminishing to 6.5 percent) of growth rate was observed as compared to saline-injected controls. Feed efficiency and body composition were reported to be unaffected by cST treatment. Subsequently, Burke et al. (1987) reported that recombinantly derived cST at a dose of 500 µg/kg body weight injected subcutaneously 3 times/day (1,500 µg/day, total) failed to stimulate rate and efficiency of gain in either male or female broiler chicks treated from 2 to 24 days of age. Despite the elevation following injection, of circulating cST by six- to sevenfold in treated birds versus controls, no effect on nitrogen retention or carcass protein and ash content were observed. Speculation was offered (Burke et al., 1987) that as a result of intensive selection for growth velocity in this species, circulating concentrations of cST do not limit growth in aves; therefore, exogenous cST administration is biologically neutral.

More recently, Vasilatos-Younken et al. (1988) reported that intravenous administration of cST episodically at a dose of 123 to 150 µg cST/kg body weight/day administered over

a 10-minute period every 90 minutes at stage of development coinciding with the nadir of basal and episodic endogenous cST secretion (8 weeks of age as reported by Vasilatos-Younkin and Zarkower, 1987) improved rate (P < 0.07) and efficiency (P < 0.01) of gain over a 21-day treatment regime of broiler-strain pullets. Modest reductions (P < 0.05 to 0.09) of indices of body lipid content were observed as compared to saline-treated controls. These changes in carcass fat content were not paralleled by a significant increase of carcass protein content. In contrast, continuous, nonepisodic infusion of cST impaired feed efficiency and had no influence on body fat indices. In several studies by other investigators, the continuous administration of cST resulted in an increase in carcass fat, with no change in growth rate (Cogburn et al., 1989b; Scanes et al., 1990). The significance of these studies is that age of birds and pattern of hormone administration are important considerations for obtaining positive responses to cST administration. In addition, if a pulsatile pattern of daily administration is required to obtain a positive response in birds, this introduces an additional complexity for the development of a delivery system for aves.

Qualitatively, avian species may appear less sensitive to, or a less likely target species for, exogenous ST treatment. However, the retardation of growth velocity and a shift toward excessive fat deposition is observed in broilers following hypophysectomy (King, 1969). Neutralization of circulating cST by passive immunization with cST antisera depressed growth velocity (Scanes et al., 1977). These studies indicate that cST is a factor influencing the physiology of nutrient partitioning; however, pharmacologic treatment of aves is at best equivocal.

Because cST exerts many of its growth and nutrient partitioning effects via IGF-I (Chapter 2), infusions or injections of this mediator have also been examined. Daily injections of 100 or 200 µg/kg body weight recombinant-derived human IGF-I, from 11 to 24 days of age, did not change the rate of gain, feed efficiency, body fat, or protein gain of broiler chicks (McGuiness and Cogburn, 1991). Likewise, the infusion of 100 µg/kg body weight/day of recombinant-derived human IGF-I did not affect rate of gain of a slow-growing brown layer strain; however, IGF-I infusion decreased abdominal body fat (Tixier-Boichard et al., 1992).

Another strategy also affecting ST-mediated physiology involves the exogenous treatment of aves with hypothalamic peptides that stimulate endogenous ST secretion. These include thyrotropin-releasing hormone (TRH, a tripeptide) and ST-releasing factor (GRF, a 44-amino acid peptide). Both TRH (Harvey et al., 1978; Scanes and Harvey, 1981) and GRF (Leung and Taylor, 1983) have been shown to stimulate cST secretion both in vitro and in vivo.

Leung et al. (1984b) found that a daily intravenous injection of TRH at doses of 1 and 10 µg/day significantly improved rate of gain in 4-week-old broiler chicks. Over the 17-day treatment regimen, rate of gain was significantly increased by an average of 12 percent compared to saline-treated control birds. Serial blood sampling for circulating cST analysis indicated that the cST secretory response to TRH diminished with treatment duration. Based on these effects, the anabolic actions of TRH were attributed to the stimulation of ST secretion and the subsequent effect on metabolism. TRH effects on triiodothyronine (T₃) and thyroxin (T₄) production were discounted because thyroid hormone treatment, per se, depresses performance and cST secretion (Leung et al., 1984a). Ingestion of TRH at a level of 10 mg/kg diet from weeks 3 to 7 of production increased daily gain by 14 percent with improved feed conversion (Cogburn et al., 1989a); however, plasma cST concentrations decreased by 33 percent.

Bolus administration of human GRF at doses of 80 and 320 µg/kg body weight/day increased the rate of body weight gain slightly in 1-to 3-week-old broiler chicks (Baile et al., 1986). Circulating cST and IGF-I were increased significantly over control values; however, neither cST nor IGF-I revealed dose responsiveness. Continuous infusion of GRF had no effect on growth rate and induced pituitary desensitization. Leung (1986) reported that doses of 0.1 and 1 µg human GRF administered intravenously once a day transitorily increased rate of gain in 4-week-old broiler chicks. Over the 14-day treatment regimen, the initial 35 percent stimulation in growth rate observed on day 3 diminished to a 9 percent advantage as compared to control chicks. Such studies are consistent with the known physiological role of GRF, but they also indicate that refractoriness to the treatment could be anticipated with pharmacologic doses.

Various analogs of epinephrine have typical cardiac activity, but several have reported striated muscle activity selectively increasing the deposition rate of lean tissue (Chapter 2). Dalrymple et al. (1984) reported that clenbuterol added to the diet of broilers between 4 and 7 weeks of age improved rate (5.1 percent) and efficiency (5 percent) of body weight gain, increased dressed carcass weight (1.1 percent), and reduced body fat content (11 to 15 percent) when used at levels of 1 mg/kg diet compared to control broilers. Abdominal fat pad weight was selectively reduced (8.5 percent) in female broilers, while no effect was apparent in males. Comparative slaughter analysis revealed that nutrient partitioning was altered and the improved performance was the result of increased carcass protein (3.6 percent) and water (2.5 percent) content combined with reduced fat content primarily in females. Minimum effective dose determined from dose-response titration suggested

1 mg clenbuterol/kg diet was necessary to evoke the nutrient partitioning effects.

Merkley and Cartwright (1989) using cimaterol at 0.25 mg/kg diet examined the effects on rate of carcass protein, water, and fat accumulation in female broilers. Cimaterol increased carcass protein and decreased carcass lipid compared to control broilers. Despite these effects on nutrient partitioning, no effects on growth performance were noted. The significant reduction in the size of the abdominal fat pad was concluded to result from a reduction of adipocyte hypertrophy and not hyperplasia. Morgan et al. (1989) fed 1 mg dietary cimaterol/kg diet to broiler chicks from 21 to 56 days of age and also observed no impact on growth rate; however, carcass fat was decreased by 20 percent. Shear force values for cooked breast meat were increased by 27 percent over unsupplemented controls, indicating increased meat toughness.

Another β-adrenergic agonist, L-644,969, has both growth-promoting and nutrient-partitioning effects in broilers (Rickes et al., 1987). A 1 to 3 percent improvement in rate of growth and a 1.5 percent improvement in feed conversion efficiency was concluded to result from a 5 to 6 percent increase in carcass protein concentration. No effect on fat deposition was apparent.

The β-adrenergic agonist, ractopamine, increases rate of gain and efficiency of gain in a dose-responsive manner during the finishing period in turkeys (Wellenreiter and Tonkinson, 1990a). Compared to controls, 44 mg ractopamine/kg diet resulted in a 17 percent increase in rate of gain and an 11 percent increase in feed conversion efficiency. Ractopamine was more effective in hens than toms. The expression of ractopamine effects was improved by increasing the dietary protein above that suggested by the NRC (1984). Ractopamine increased carcass weight, dressing percentage, and leg and thigh as percent of the carcass and decreased the abdominal fat pad as percent of the lean weight (Wellenreiter and Tonkinson, 1990b). In this study, shear values for the pectoralis major muscle were not affected by ractopamine.

Theoretically, passive or active immunization of animals against somatostatin would remove the intrinsic negative control over cST secretion as well as thyroid stimulating hormone (TSH) and insulin (Vale et al., 1974). As cST, TSH, and insulin concentrations are all major positive hormonal stimuli for IGF-I and growth, immunological approaches have focused on somatostatin.

Lam et al. (1986) reported that both 4- and 8-week-old broilers treated with sheep anti-somatostatin serum had a marked increase in circulating T₃ and T₄ concentrations compared to chicks receiving normal sheep serum. The effect occurred within 10 minutes of treatment and was sustained about 5 hours. Similarly, passive immunization against somatostatin evoked a marked increase of circulating cST (ninefold more than baseline) within 10 minutes of treatment and remained elevated approximately 90 minutes (Harvey et al., 1986). The effects of prolonged treatment and the consequences on growth and nutrient partitioning have not been reported.

Restriction of feed intake to maintenance energy 6 to 12 days posthatch followed by release to ad libitum intake results in true compensatory growth (fractional growth rate accelerated compared to ad libitum fed controls; Plavnik et al., 1986). During the 8-week production period, restricted chicks weighed slightly more than control broilers resulting in a significant improvement in daily gain. Feed efficiency was improved by 10 percent over the production period in both male and female birds. Total carcass fat was reduced by 17 percent, and the abdominal fat pad size was reduced 30 percent. McMurtry et al. (1988) reported that feed restriction early in life delays the peak of circulating cST from day 12 posthatch noted in control chicks until day 42, coincident with the maximum compensatory growth period of restricted birds. Both circulating T₃ and T₄ concentrations were reduced in restricted chicks. Manipulation of feed intake pattern in broilers was suggested (McMurtry et al., 1988) to result from a delay of physiological maturity, such that intrinsic nutrient-partitioning priority of birds during the finishing phase of growth, typified by fat accretion, is delayed, thereby prolonging the period of maximum protein accretion.

An attractive strategy for manipulating broiler growth is the injection of hormones or pharmacologically active substances into the egg to elicit changes in embryonic development that will result in desirable changes in growth characteristics of the chick. Several rationales are being pursued, although little has been published. First, injected substances may speed the growth of the developing embryo so that the hatched chick is bigger. Because there is considerable unused yolk and albumen in newly hatched chicks, a large pool of additional nutrients is available for growth of the embryo prior to hatching. Following hatching, it is presumed that a chick starting at a greater initial weight and growing at a normal fractional rate will have a faster absolute rate of growth (g/day). The second rationale is to deliver an agent to the embryo at a critical stage of development so that the course of development is changed. Two approaches are suggested: (1) increase the number of cycles of division of embryonic myoblasts before fusion so that additional myofibers will be available for hypertrophy in the growing chick; (2) decrease the number of progenitor cells for adipose cells

so that fewer preadipocytes are available for hypertrophy and hyperplasia in the growing chick.

Egg injection systems are becoming commercially available and will facilitate the delivery of pharmacological agents to the embryo. This approach is particularly attractive because manipulation of the embryo is considerably more feasible than injections or implants in the growing chick and could possibly require only one or several injections. Hargis et al. (1989) have administered a single injection of ST into the albumen of 11-day- old eggs and observed an 11 percent increase in body weight of 7-week-old male chicks. The increased growth rate was accompanied by improved feed conversion. Although carcass composition was not altered, in ovo ST administration resulted in fewer adipocytes per gram of adipose tissue, but individual cells had greater volume.

A stated requirement for a given nutrient represents a single point along a dose-response curve that when applied to a well-characterized animal population can, with some reliability, achieve predictable growth performance or targeted composition of body weight gain. Application of metabolic modifiers to a target species, such as broilers, represents a unique challenge to nutritionists because such technologies are not well characterized with respect to the population in general. More important, heterogeneity of the response, whether it be a growth enhancement with no effect on composition or a true change in the priority of nutrient partitioning, confounds nutrient characterization. Therefore, only deductive reasoning applied primarily to protein (amino acid) and energy nutriture can be offered with the rationale that this suggested standardized formulation would not limit biological responsiveness for those strategies that affect nutrient partitioning.

Traditionally, nutrient requirements for broilers have been determined empirically using data from feeding trials. The lowest level of a nutrient that results in optimal growth, feed efficiency, and carcass composition is considered to be the requirement. Because of the short growing period (7 weeks) and relative ease of procuring and conducting research with the growing broiler, accurate requirement data have been developed for many of the nutrients. Historically, dietary recommendations made by the NRC have been divided into three growth periods; starter (0 to 3 weeks), grower (3 to 6 weeks), and finisher (6 to 8 weeks). Because the bird's nutrient requirements do not abruptly change at each of these periods, most commercial broiler producers adjust the nutrient levels more frequently, often 5 or 6 times during the 7-week growth period of a broiler, to keep the diet ''least-cost." Multiple diets are necessary for least-cost formulations because dietary protein is expensive and the amino acid requirement (as percentage of the diet or percentage of the calories) decreases with age. From a research perspective, the three dietary divisions suggested by the NRC (1984) are adequate to provide a diet reasonably close to those actually used commercially, yet meeting or exceeding all of the bird's nutrient requirements.

Assuming that the objective of experiments concerning metabolic modifiers is to accelerate the rate of lean tissue growth, amino acid requirements will generally be increased compared to current NRC requirements. Over-formulation of dietary amino acids to ensure adequate supply to support a greater rate of protein accretion may be prudent in some situations because excess amino acids will be deaminated and used for energy. However, it is likely that the action of a metabolic modifier may interact with dietary protein level and that effects seen with unrealistically high levels of protein may be considerably different than those observed with "least-cost" diet formulations supplying only enough protein to meet the amino acid requirement. The influence of high dietary protein to energy ratios on circulating hormone levels and carcass composition is well documented (Rosebrough and Steele, 1985). Thus, our goal should be to use experimental diets with additional nutrient fortification to meet the anticipated increased growth rates and increased percent lean in the carcass without supplying surfeit levels of amino acids.

Ideally, mechanistic models should be used to predict changes in nutrient requirements that occur when metabolic modifiers are used. Mechanistic models are based on known metabolic pathways and a quantitative description of their regulation. Thus, mechanistic models derive nutrient requirements from an accounting of their metabolic sources. Metabolic modifiers would drive changes in intermediary metabolism and thus the model would predict new nutrient requirements. Unfortunately, no mechanistic models have been written for poultry.

Most current models for predicting nutrient requirements of animals with variable rates of productivity are based on a trial-and-error approach using input-output relationships (deductive or factorial approach). The input required to support a given rate of growth at any carcass composition can be estimated if maintenance needs and the partial efficiency of nutrient use for productive processes are known. These conceptual relationships form the basis for several models of nutrient requirements of ruminants. This type of prediction has been largely neglected in growing poultry because it is easy to obtain accurate nutritional requirements by conducting growth trials. Thus, there is a lack of sound experimental data on the amino acid and energy requirements needed for maintenance, lean tissue accretion, and adipose tissue accretion. Using general assumptions for maintenance requirements relative to body size, Scott et al.

(1982) used output-based models to predict energy and protein requirements for growing broilers or White Leghorn pullets. Hurwitz and co-workers (1978, 1980) developed similar models to predict amino acid requirements of broilers. The broiler model predicts amino acid requirements (mg/kcal) considerably lower than those suggested by the NRC (1984) based on growth trials, particularly for tryptophan, lysine, and methionine. Because diets formulated to levels predicted by the model resulted in a large increase in abdominal fat, it can be concluded that the model underestimated the true requirement for at least some of the amino acids. A modeling approach is also limited by lack of accurate information on the changes with age of the amino acid composition of the carcass and the maintenance need for energy and amino acids. Until more accurate information on these parameters across various ages of the growing broiler chick are experimentally determined, input-output based models have limited use for accurate prediction of amino acid requirements of broilers. Input-output models have been useful for demonstrating relative changes in requirements as growth rate and carcass composition change. They predict that, at a given rate of growth, improved composition of gain increases amino acid requirements but have little effect on energy requirements. Thus, amino acid requirements expressed as milligrams per kilocalorie increase with improved composition of gain. Input-output models also predict that increasing the rate of growth by 10 percent increases amino acid requirements 10-fold more than decreasing the carcass fat by 10 percent at a slaughter weight of 2.5 kg.

Another way to model changes in nutrient requirements caused by metabolic modifiers is to use empirically obtained nutrient requirements over a range of growth rates to predict requirements at new rates. Because of the ease of conducting experiments with broilers, the requirements for energy, lysine, and methionine are known with reasonable accuracy during the commercially important first 8 weeks of growth. Although prediction based on empirical relationships is not easily amenable to predicting requirements at a variety of carcass compositions, this procedure has a considerably better data base and requires fewer assumptions than the input-output modeling method.

As broiler chicks get older, their amino acid requirements decrease relative to energy. This is partly the result of a decreased fractional growth rate and partly the result of an increased rate of fat relative to protein accretion. Regression of fractional growth rates calculated from data on male broilers (National Research Council, 1984) against amino acid requirements in grams per kilocalorie results in linear relationships (Table 6-1). Expression of amino acid and mineral requirements relative to energy density of the diet is useful because the energy level of the diet is important in controlling feed intake. It is assumed that without a change in the composition of gain, increased growth rates result in increased energy needs and thus increased feed intake. Use of these relationships to predict amino acid requirements for broilers growing at 10, 20, and 30 percent faster rates is shown in Table 6-2. A similar approach can be used to predict requirements for macrominerals at various fractional growth rates. Recommended diets based on corn and soybean meal for broiler chicks used in experiments with metabolic modifiers are shown in Table 6-3. Equations are based on the assumption that chicks are raised at thermoneutral environmental temperatures.

The use of empirically determined requirements from data across the growth period to predict requirements at augmented growth rates relies on several assumptions. First, this empirical approach assumes that the carcass composition of broilers given a metabolic modifier is similar to that of normal broilers at that part of their growth curve where they have the same fractional growth rate. Because normal broilers increase fat to lean ratios as they get older and their growth rates decrease, it follows that the regression approach assumes that a broiler that reaches market weight at a higher fractional growth rate is leaner than a normal broiler

that reaches the same market weight at an older age. In practical terms, the amino acid requirements predicted by the regression equations assume that the carcass composition of female broilers growing at a 20 percent greater fractional rate have 18 percent less fat at the same market weight (2.2 kg) based on the fat accretion data of Hood (1982). As discussed above, many metabolic modifiers cause both a decrease in lipid accretion and an increase in protein accretion so that this assumption may be generally correct.

Second, use of empirically derived requirements assumes that the relationship between amino acid requirements and growth rates is linear. Although this assumption is valid for birds with normal growth rates, it is not known whether the relationship holds at augmented rates.

Third, it is assumed that metabolic modifiers inducing augmented fractional growth rates do not change the proportion of amino acids required for maintenance relative to protein accretion. Certainly this assumption must be considered for each specific metabolic modifier because some are known to markedly affect the rate of amino acid catabolism (Chapter 2). Some anabolic hormones decrease rates of amino acid deamination and use for processes other than protein accretion. Consequently, use of values in Table 6-2 would tend to overestimate the amino acid requirement.

Few researchers have reported details on diets fed to experimental chicks exposed to metabolic modifiers. With a few notable exceptions, it is generally not clear if these diets meet accepted feeding standards such as NRC guidelines. Thus, it cannot be determined if adequate dietary amino acids were provided to permit improved growth rate or increased lean composition. Investigators are encouraged to publish descriptions of diets used in research on metabolic modifiers and use diets similar to those described here to assure that the diet does not limit the physiological expression of responses.

Data demonstrating successful use of metabolic modifiers with poultry have not been sufficiently encouraging to stimulate development of products for market by pharmaceutical and allied industries. Only recently have recombinant chicken hormones become available, and there is a dearth of work with recombinant turkey hormones. It has also been speculated that the high reproductive capacity of poultry has permitted traditional breeding programs to optimize endogenous levels of anabolic hormones. Consequently, supplementation of additional hormone results in little augmentation of response. Alternately, it may be that our appreciation for the regulation of growth and lean tissue accretion gained using mammals is not relevant to poultry, and other classes of metabolic modifiers should be appraised for efficacy. In either case it is important that diets be used that permit the expression of increased growth and/or improved lean tissue accretion. Current knowledge of nutrient requirements versus growth rate of broilers indicate that diets that meet current NRC nutrient levels are not adequate for research on metabolic modifiers. Diets that will permit the expression of accelerated growth and lean tissue accretion require higher amino acid and macromineral levels in proportion to metabolizable energy content of the diet. Failure to use appropriate diets may prevent the realization of augmented gain and improved lean tissue accretion in experiments with metabolic modifiers.