2005 Poultry Science Association, Inc.

Relative Effectiveness of Methionine Sources in Diets for Broiler Chickens D. Hoehler,*,1 A. Lemme,† S. K. Jensen,‡ and S. L. Vieira§ *Degussa Corporation, 1701 Barrett Lakes Blvd., Suite 340, Kennesaw, Georgia 30144; †Degussa AG, Feed Additives, Rodenbacher Chaussee 4, 63457 Hanau, Germany; ‡Department of Animal Nutrition and Physiology, Research Centre Foulum, PO Box 50, DK-8830 Tjele, Denmark; and §Universidade Federal do Rio Grande do Sul, Departamento de Zootecnia, Av. Bento Goncalves, 7712, Porto Alegre, RS - 91540-000, Brazil

Primary Audience: Researchers, Nutritionists, Extension Specialists, Broiler Producers SUMMARY There is an ongoing discussion regarding the relative effectiveness of the hydroxy analog of methionine (liquid MHA-FA) relative to DL-methionine (DLM). Five experiments were carried out under local conditions in 4 different countries to further test and determine the relative effectiveness of liquid MHA-FA. Additionally, multiexponential regression was used to determine if it is the proper mathematical model for estimation of Met relative effectiveness. The trials were carried out at the Research Center Foulum, Perdue Farms Inc., a major Mexican integrator, the University of Arkansas, and at the University of Rio Grande do Sul. Basal diets in all trials were formulated to be deficient in Met but adequate in all other nutrients and energy. Broilers performed well in each experiment; significant responses showed the Met deficiency of the basal diets. Relative effectiveness estimates of diluted DLM (65%) was 63% on average (n = 6, based on 3 trials), confirming that simultaneous regression analysis represents the proper mathematical model for comparative nutritional purposes. Relative effectiveness of liquid MHA-FA relative to DLM was 64% (n = 11, based on all 5 trials) on average for weight gain, feed conversion, and breast meat yield. Key words: broiler, amino acid, methionine source, bioefficacy, exponential regression 2005 J. Appl. Poult. Res. 14:679–693

DESCRIPTION OF PROBLEM In commercial poultry diets, Met is usually the first limiting amino acid and is commonly supplemented as dry DL-Met (DLM; 99% pure) or as liquid DL-Met hydroxy analog-free acid (MHA-FA, containing 88% of active substance). Knowing the relative nutritive value of liquid MHA-FA compared with DLM is an important precondition to cost-effective purchasing, feed formulation, and animal production. There is an ongoing discussion in the liter1

ature regarding the relative bioefficacy of liquid MHA-FA relative to DLM [1, 2, 3, 4, 5, 6, 7, 8]. Recent work by Lemme et al. [9] reported an average relative effectiveness of liquid MHA-FA compared with DLM of 62% based on weight gain, feed conversion, breast meat yield, and carcass yield in broiler chickens. This result was derived from 2 extensive trials carried out in Canada and Australia. These results also suggested that simultaneous exponential regression analysis, as described by Littell et al. [10], is a valid statistical approach for deter-

To whom correspondence should be addressed: [email protected].

JAPR: Research Report

680 mination of the relative bioefficacy of Met sources. Additionally, researchers in The Netherlands [11] conducted an independent, extensive literature study and evaluation of previous papers on this topic and reported that the average bioefficacy of liquid MHA-FA compared with DLM in broilers was 68% on a weight basis or 77% on an equimolar basis. These figures are based on weight gain and feed conversion of 18 broiler trials. The primary objective of the present study was to validate the former review [11] on Met source efficacy in 5 experiments conducted at different institutions. Relative efficacy of Met sources was evaluated in all 5 experiments, and the suitability of the exponential regression method for bioefficacy determination was evaluated in 3 experiments.

MATERIALS AND METHODS Trial 1 was conducted at the Research Center Foulum (Tjele, Denmark). We assigned 2,880 male, 1-d-old Ross 208 broilers [12] to 16 dietary treatments for 42 d. Each treatment was replicated 6 times with 30 birds per replicate. Treatments were composed of a basal wheat-soybean meal-pea diet (Table 1) and 3 series of diets containing graded levels (0.04, 0.08, 0.12, 0.16, and 0.20%) of DLM, diluted DLM (65%), or liquid MHA-FA. For each treatment, a starter was fed from d 1 to 21, and a grower diet was fed from d 22 to 42. Mash feed and water were offered ad libitum. Basal diets were formulated to be adequate for energy and all nutrients [13] except Met + Cys, which were 0.59 and 0.52% in the starter and grower diets, respectively. Aliquots of the basal diets were then supplemented with graded levels of the Met sources. Glucose was added to dilute DLM to make the DLM 65% product. Birds were allocated to floor pens of 1.7 m2, which had wheat straw as bedding. After 3 d, temperature was gradually decreased from 33 to 21°C. There was continuous light throughout the experiment. Body weights and feed consumption were recorded for the 1-to-21 and 21-to-42-d periods. Subsequently weight gain and mortality-corrected feed conversion were calculated. In trial 2 at Perdue Farms Inc. (Salisbury, MD) 4,500 male, 1-d-old Perdue 5632 broilers [14] were distributed among 9 dietary treat-

TABLE 1. Ingredients, energy and nutrient content of the basal starter and grower diets, trial 1 (Research Centre Foulum, Tjele, Denmark) Ingredient, % Wheat Soybean meal Peas Soybean oil Animal fat Lys (blend 40 %) Thr (blend 50 %) Calcium carbonate Dicalcium phosphate Salt (NaCl) Sodium bicarbonate Vitamin-mineral premix Choline chloride, 50 % Xylanase Total energy1 and nutrient content ME, kcal/kg ME, MJ/kg Crude protein, % Met, % Met + Cys, % Lys, % Thr, % Arg, %

Starter 1–21 d

Grower 22–42 d

55.40 22.50 12.50 3.00 2.50 0.59 0.16 1.20 1.30 0.30 0.20 0.20 0.05 0.10

54.70 17.00 16.00 4.00 4.00 0.67 0.28 1.20 1.30 0.30 0.20 0.20 0.05 0.10

2,950 12.5 19.6 0.27 0.59 1.13 0.73 1.31

3,080 12.9 18.0 0.23 0.52 1.06 0.72 1.19

1

Calculated.

ments. Five floor pens with 100 birds each were allotted to each treatment. Nine experimental diets per phase were produced from a basal diet, which was deficient in Met + Cys and 8 diets with 4 graded levels (0.06, 0.12, 0.18, and 0.24%) of DLM or liquid MHA-FA. Birds received the starter diets from d 1 to 18, the grower 1 diets from d 19 to 28, and the grower 2 diets from d 29 to 39. Pelleted feed and water were offered ad libitum. The basal diets were based on corn, soybean meal, meat and bone meal, and bakery meal and were formulated to be adequate for energy and all nutrients [13] except for Met + Cys (Table 2). The basal Met + Cys contents were 0.68, 0.65, and 0.62% in the starter, grower 1, and grower 2 diets, respectively. Body weights as well as feed consumption were recorded at d 19 and at the termination of the experiment at d 39. Pens were checked daily for mortality. Weight gain and mortality-corrected feed conversion were then calculated. In trial 3 at a major Mexican integrator, 4,680 male 1-d-old Hybro broilers [15] were

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS

681

TABLE 2. Ingredients and energy and nutrient contents of the basal starter and grower diets, trial 2 (Perdue Farms, Inc., Salisbury, MD) Ingredient, % Corn Soybean meal Meat and bone meal Bakery meal Soybean hulls Other1 Total energy2 and nutrient content ME, kcal/kg ME, MJ/kg Crude protein, % Met, % Met + Cys, % Lys, % Thr, % Arg, %

Starter 1–18 d

Grower 1 19–28 d

Grower 2 29–39 d

56.84 20.45 8.00 6.00 5.00 3.71

59.29 18.70 7.39 5.00 5.00 4.62

61.15 17.50 6.34 5.00 5.00 5.01

3,110 13.0 20.0 0.35 0.68 1.20 0.75 1.34

3,170 13.3 18.8 0.33 0.65 1.10 0.71 1.25

3,200 13.4 17.7 0.32 0.62 1.04 0.67 1.17

1

Other: Lys, minerals, trace elements, vitamins, and phytase; identical for the different dietary treatments and according to Perdue’s commercial guidelines. 2 Calculated.

assigned to 13 dietary treatments for 42 d. Five floor pens with 72 birds each were allotted to each treatment. Treatments were composed of a basal sorghum-soy diet and 3 series of diets containing graded levels (0.06, 0.12, 0.18, and 0.24%) of DLM, diluted DLM (65%), or liquid MHA-FA. Birds received the starter diets from d 1 to 21, the grower diets from d 22 to 32, and the finisher diets from d 33 to 42, respectively. Pelleted feed and water were offered ad libitum. Basal diets were mainly based on sorghum, soybean meal, and meat and bone meal and were formulated to be adequate for energy and all nutrients [13] except for Met + Cys (Table 3). The Met + Cys contents were 0.65, 0.60, and 0.55% in the starter, grower, and finisher diets, respectively. Aliquots of the basal diets were then supplemented with graded levels of the respective Met sources. To obtain DLM (65%), finely ground sorghum was used to dilute DLM. Body weights as well as feed consumption were recorded at d 1, 21, 32 and 42 (termination of the experiment). Pens were checked daily for mortality. Dates and body weights of dead broilers were recorded. Weight gain and mortality-corrected feed conversion were then calculated. In trial 4 at the University of Arkansas (Fayetteville, AR), 1-d-old male Ross 308 broilers [16] were assigned to 16 dietary treatments

from d 7 to 35. Treatments were composed of a basal corn-soy diet and 3 series of diets containing graded levels (0.03, 0.06, 0.09, 0.12, and 0.15%) of DLM, diluted DLM (65%), or liquid MHA-FA. A common starter diet containing 23% CP was fed from d 1 to 6. On d 7, birds were weighed, and the light and heavy birds were removed from the population. A total of 576 broilers were distributed to 96 floor pens containing 6 birds each. Six pens were assigned to each of the 16 dietary treatments in a complete block design to achieve equal body weight distribution between treatments. The basal diet was formulated to be adequate for energy and all nutrients [13] except for Met (0.30) and Met + Cys (0.62%, Table 4). The highest DLM supplementation level was 0.15% in diet 6, which amounted to a maximum Met + Cys level of 0.79% in that treatment. Diluted DLM (65%) was prepared by blending DLM (99%) with finely ground corn meal. Pelleted feed and water were offered ad libitum. Body weights as well as feed consumption were recorded at d 7 and 35. Mortality was recorded, and feed conversion was corrected for mortality. Trial 5 was conducted at the Federal University of Rio Grande do Sul in Brazil using 1-d-old Ross 308 broilers [17]. From hatching to 7 d of age, 2,730 chicks were fed a commer-

JAPR: Research Report

682

TABLE 3. Ingredients and energy and nutrient contents of the basal starter, grower, and finisher diets, trial 3 (Mexican Integrator) Starter 1–21 d

Grower 22–32 d

Finisher 33–42 d

63.00 27.90 5.00 1.85 0.41 0.68 0.08 0.35 0.31 0.13 0.29

67.00 23.10 5.00 2.82 0.39 0.50 0.08 0.10 0.26 0.10 0.65

70.00 18.70 5.00 4.13 0.47 0.45 0.08 0.03 0.23 0.08 0.83

3,000 12.6 21.2 0.32 0.65 1.26 0.80 1.42

3,100 13.0 19.3 0.30 0.60 1.12 0.72 1.27

3,200 13.4 17.5 0.27 0.55 1.04 0.65 1.12

Ingredient, % Sorghum Soybean meal Meat and bone meal Soybean oil Biolys1 Calcium carbonate Sodium bicarbonate Ortophosphate NaCl Choline chloride 50 % Vitamin-mineral premix Total energy2 and nutrient content ME, kcal/kg ME, MJ/kg Crude protein, % Met, % Met + Cys, % Lys, % Thr, % Arg, % 1

Contained at least 50.7% L-Lys. Calculated.

2

cial mash starter diet formulated to contain 23% CP and 2,950 kcal of ME/kg. At 7 d of age birds were weighed, and 5 chicks per pen constituting TABLE 4. Ingredients and energy and nutrient contents of the basal diet fed from d 7 to 35, trial 4 (University of Arkansas, Fayetteville, AR) Ingredient

%

Corn Soybean meal Meat and bone meal Poultry oil L-LysⴢHCl L-Thr Calcium carbonate NaCl, iodized Sodium bicarbonate Vitamin-mineral premix Total energy1 and nutrient content ME, kcal/kg ME, MJ/kg Crude protein, % Met, % Met + Cys, % Lys, % Thr, % Arg, %

67.97 21.51 7.40 1.24 0.24 0.09 0.45 0.30 0.20 0.60

1

Calculated.

3,080 12.9 20.2 0.30 0.62 1.20 0.80 1.29

mean pen weight extremes were removed to improve the within-pen weight uniformity. From d 7 to 40, chicks were assigned to 13 dietary treatments housed in floor pens receiving corn/soybean-based diets supplemented with DLM (0.030, 0.060, 0.100, 0.140, 0.190, and 0.240%) or equimolar levels of liquid MHA-FA (0.034, 0.068, 0.114, 0.159, 0.216, and 0.273%) in the starter (d 7 to 21) and grower (d 22 to 40) periods (Table 5). Each dietary treatment consisted of 6 replicates with 35 birds per pen. Basal diets were formulated to be deficient in Met + Cys but adequate in all other nutrients and energy [13]. Feed and water were provided ad libitum. Overall management was as in a commercial Brazilian operation with normal environmental conditions for temperature and ventilation. Light was provided 24 h/d during the first week; afterward it followed the natural daylight regimen. All birds were weighed as a group per pen at the beginning of the experiment and then at 7, 21, and 40 d of age. Feed consumption was measured, and then mortality-corrected feed conversion was calculated. At the end of the experiment, 6 birds per pen with body

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS TABLE 5. Ingredients and energy and nutrient contents of the basal starter and grower diets, trial 5 (Federal University of Rio Grande do Sul, Porto Allegro, Brazil) Ingredient, (%) Corn Soybean meal Soybean oil L-Lys-HCl L-Thr Dicalcium phosphate Limestone Cornstarch Mineral-vitamin premix NaCl Sodium bicarbonate Choline chloride, 60% Total energy1 and nutrient content, % ME, kcal/kg ME, MJ/kg Crude protein Met Met + Cys Lys Thr Arg

Starter 7–21 d

Grower 22–40 d

58.06 32.32 3.90 0.23 0.09 1.94 1.32 1.00 0.50 0.34 0.24 0.06

59.15 30.22 5.20 0.23 0.04 1.89 1.15 1.00 0.50 0.23 0.33 0.06

3,100 13.0 20.0 0.31 0.65 1.24 0.86 1.30

3,200 13.4 19.0 0.29 0.62 1.18 0.79 1.23

1

Calculated.

weights representative of the pen average were selected and killed, defeathered, and eviscerated. Carcasses were chilled in slush ice for 3 h. The carcasses were processed in commercial parts: deboned breast, wings, thighs, and drumsticks. Weights of cuts were related to the carcass weight and expressed as percentage of carcass. Analyses verified accurate feed production and DLM, diluted DLM (65%), or liquid MHAFA supplementation. All experiments were conducted in accordance with local animal care guides. Basal diets were formulated to be deficient in Met but adequate in all other nutrients [13]. Protein-bound and supplemented amino acids of ingredients and experimental diets were determined according to Llames and Fontaine [18]. Liquid MHA-FA was analyzed in the diets by using the method described by Naumann et al. [19]. Diets for all trials were formulated to meet or exceed the essential amino acids, energy, vitamin, and mineral needs of broiler chickens except for Met + Cys [13]. Data were evaluated by ANOVA including comparison of means and simultaneous expo-

683

nential regression to estimate the relative effectiveness of diluted DLM (65%, only trials 1, 3, and 4) and liquid MHA-FA compared with DLM (99%). Pen means were considered the experimental unit for all statistical analyses. A nonlinear exponential model was used to estimate the efficacy of liquid MHA-FA and diluted DLM (65%) relative to pure DLM [9, 10]. The general linear model procedure (PROC GLM) in the SAS/STAT software [20] was applied fitting the following nonlinear equation: y = a + b × (1 − e(c1×x1+c2×x2+c3×x3)) where y = performance criterion (weight gain, feed conversion, or breast meat yield), a = intercept (bird performance with basal diet), b = asymptotic response, a + b = common asymptote (maximum performance level), c1 = steepness coefficient for pure DLM (99%), c2 = steepness coefficient for diluted DLM (65%), c3 = steepness coefficient for liquid MHA-FA, and x1, x2, x3 = dietary level of pure DLM (99%), diluted DLM (65%), and liquid MHAFA, respectively. According to Littell et al. [10], bioefficacy values for diluted DLM (65%) and liquid MHA-FA relative to DLM (99%) are given by the ratios of regression coefficients; C2/C1 = diluted DLM and C3/C1 = liquid MHA-FA. Diluted DLM (65%) was added at the same levels as pure DLM in trials 1, 3 and 4. The supplemented levels were confirmed by analysis. In these trials, a biological effectiveness of 65% could be assumed a priori for diluted DLM (65%) relative to undiluted (DLM 99%). Thus, these treatments could be regarded as an internal standard to check the validity of exponential regression analysis. The degree of dilution was chosen because most recent literature suggests an average relative effectiveness of liquid MHA-FA compared with DLM of about 65% (Lemme et al. [9], 62%, based on 2 trials; Jansman et al. [11], 68%, based on the review of 18 trials). Recently, there has been some discussion at scientific meetings regarding the proper statistical methodology to compare Met sources. Among numerous others in the scientific literature, Littell et al. [10] and Jansman et al. [11] agree that the results of such comparison experiments are analyzed efficiently by lin-

JAPR: Research Report

684

TABLE 6. Performance of broiler chickens fed graded levels of DL-Met, diluted DL-Met (65%), or liquid hydroxy analog of Met (MHA-FA) from 1 to 42 d of age (Trial 1, Research Centre Foulum, Denmark) Treatment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Met source Control DLM DLM DLM DLM DLM Diluted DLM (65 Diluted DLM (65 Diluted DLM (65 Diluted DLM (65 Diluted DLM (65 Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA

Addition of product (%) 0 0.04 0.08 0.12 0.16 0.20 0.04 0.08 0.12 0.16 0.20 0.04 0.08 0.12 0.16 0.20

%) %) %) %) %)

Weight gain (g; mean ± SD) 1,130 1,715 2,084 2,248 2,234 2,225 1,582 1,902 2,058 2,158 2,218 1,591 1,923 2,005 2,086 2,216

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

82f 110ef 73abcd 47a 7ab 65ab 54f 105de 63abcd 60abc 89ab 105f 73cde 62bcd 83abcd 121ab

Feed per gain (kg/kg; mean ± SD) 2.145 2.002 1.934 1.888 1.869 1.829 2.085 1.995 1.918 1.923 1.874 2.023 1.945 1.952 1.917 1.877

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.035a 0.068bcd 0.050cde 0.063cde 0.033de 0.044e 0.057ab 0.044bcd 0.034cde 0.053cde 0.026de 0.068abc 0.052cde 0.060bcde 0.044cde 0.051de

Means ± SD in a column with no common superscript differ significantly (P < 0.05).

a–f

ear (slope-ratio) or nonlinear (exponential) models.

RESULTS AND DISCUSSION Broilers performed well in each experiment. Total mortalities were 2.3, 3.9, 4.8, 3.5, and 1.4% in the experimental periods of trials 1 to 5, respectively, and were relatively low and not affected by dietary treatments. Trial 1 Broilers responded very well to all supplemental Met sources. Due to the low Met + Cys content in the basal diet, trial 1 had the greatest performance response from the basal diet to

the highest Met-supplemented diet of 97% for weight gain, from 1,130 to 2,225 g, and of 15% for feed conversion, from 2.145 to 1.829 kg/kg (Table 6). The strong responses on performance confirmed that the basal diets were Met + Cys deficient, thus, weight gain achieved by the highest inclusion level was almost twice as high as that achieved by the basal diet. Variability was low and therefore performance of the basal treatment group was significantly different from that of higher inclusion levels. With respect to weight gain, maximum performance was already achieved at 0.12% inclusion of DLM, whereas 0.20% was needed for the treatments with diluted DLM (65%) and liquid MHA-FA. Minimum feed conversion of 1.829

TABLE 7. Performance of broiler chickens fed graded levels of DL-Met or liquid hydroxy analog of Met (MHA-FA) from 1 to 39 d of age (trial 2, Perdue Farms Inc., Salisbury, MD) Treatment 1 2 3 4 5 6 7 8 9

Met source

Addition of product (%)

Control DLM DLM DLM DLM Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA

0 0.06 0.12 0.18 0.24 0.06 0.12 0.18 0.24

Body weight (g; mean ± SD) 1,717 1,812 1,820 1,890 1,895 1,786 1,827 1,838 1,871

± ± ± ± ± ± ± ± ±

75b 47ab 60ab 11a 56a 43ab 66ab 49ab 34a

Means ± SD in a column with no common superscript differ significantly (P < 0.05).

a,b

Feed per gain (kg/kg; mean ± SD) 1.722 1.696 1.662 1.635 1.628 1.703 1.684 1.678 1.639

± ± ± ± ± ± ± ± ±

0.030a 0.020ab 0.038ab 0.014b 0.024b 0.018ab 0.046ab 0.027ab 0.026b

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS

685

FIGURE 1. Weight gain and feed conversion of broilers fed increasing levels of DL-Met (DLM, 99%), diluted DLM (65%), or liquid methionine hydroxy analog (88%, liquid MHA-FA) from d 1 to 42 (trial 1, Research Centre Foulum, Tjele, Denmark). Values in parentheses indicate 95% confidence intervals. *Significantly less than 88%; P < 0.05.

of birds fed the 0.20% DLM diet was not achieved by feeding either diluted DLM (65%) or liquid MHA-FA. As shown in Figure 1, the response data for both performance criteria followed a nonlinear trend corresponding to the law of diminishing returns. The strong nonlinear response, the low variability, and a sufficient number of replicates and birds per replicate allowed for simultaneous exponential regression to determine the relative effectiveness of diluted DLM (65%) and liquid MHA-FA. According to the regression analysis, liquid MHA-FA was only 64 and 67% as efficient as DLM regarding weight gain and feed conver-

sion, respectively. Both results were significantly lower than 88%, which is the content of active substance in liquid MHA-FA. For diluted DLM (65%), the estimated biological efficiency was 67% for weight gain and 59% for feed conversion, which represents a good agreement with the expected result of 65% for this product. Trial 2 The statistically significant responses in weight gain and feed per gain between the control treatment and the treatments with highest inclusion rates confirmed that the basal diets

JAPR: Research Report

686

FIGURE 2. Weight gain and feed conversion of broilers fed increasing levels of DL-Methionine (DLM, 99%) or liquid methionine hydroxy analog (88%, liquid MHA-FA) from d 1 to 39 (trial 2, Perdue Farms, Inc., Salisbury, MD). Values in parentheses indicate 95% confidence intervals. Feed conversion ratio was standardized to a final weight of 4.05 lb.

were Met + Cys deficient (Table 7), although performance responses of weight gain (+17%) and feed conversion (−10%) were not as drastic as in trial 1. The excellent feed conversion of only 1.628 kg/kg in treatment 5 with the highest DLM supplementation can be explained by the relatively short trial period of 39 d. The response data for both performance criteria followed a nonlinear trend allowing for multiexponential regression analysis (Figure 2). Liquid MHA-FA was only 70 and 66% as efficient as DLM regarding weight gain and feed conversion, respectively. Both figures were nonsig-

nificant, but especially the 95% confidence interval for feed conversion (liquid MHA-FA: 41 to 90%) was relatively low. Trial 3 The statistically significant responses of weight gain and feed per gain between the control treatment and the treatments with highest DLM inclusion rates of 17 and 10%, respectively, confirmed that the basal diets were Met + Cys deficient (Table 8). The response data for both performance criteria followed a nonlinear trend allowing for multiexponential regression

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS

687

TABLE 8. Performance of broiler chickens fed graded levels of DL-Met, diluted DL-Met (65%), or liquid hydroxy analog of Met (MHA-FA) from 1 to 42 d of age (trial 3, Mexican Integrator, Mexico) Treatment 1 2 3 4 5 6 7 8 9 10 11 12 13

Met source Control DLM DLM DLM DLM Diluted DLM (65 Diluted DLM (65 Diluted DLM (65 Diluted DLM (65 Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA

Addition of product (%)

%) %) %) %)

0 0.06 0.12 0.18 0.24 0.06 0.12 0.18 0.24 0.06 0.12 0.18 0.24

Weight gain (g; mean ± SD) 1,777 2,015 2,021 2,086 2,083 1,939 2,038 2,065 2,078 1,952 2,001 2,068 2,073

± ± ± ± ± ± ± ± ± ± ± ± ±

116c 33ab 28ab 26a 51a 49b 30ab 32ab 31a 50ab 20ab 38ab 55ab

Feed per gain (kg/kg; mean ± SD) 1.921 1.805 1.760 1.732 1.724 1.820 1.764 1.750 1.747 1.815 1.797 1.745 1.728

± ± ± ± ± ± ± ± ± ± ± ± ±

0.094a 0.032b 0.023b 0.012b 0.026b 0.039b 0.028b 0.018b 0.021b 0.018b 0.024b 0.034b 0.016b

Means ± SD in a column with no common superscript differ significantly (P < 0.05).

a–c

analysis (Figure 3). Accordingly, liquid MHAFA was 63 and 73% as efficient as DLM regarding weight gain and feed conversion, respectively. For diluted DLM (65%), the estimated biological efficiency relative to pure DLM was 69% for weight gain and 79% for feed per gain (Figure 3). Although the figure for weight gain was within the expected range, the figure for feed conversion was pretty high which was similar to the value for liquid MHA-FA. Trial 4 Broilers performed well; the performance response verified that the basal diet was deficient in Met + Cys. Weight gain achieved at the highest DLM inclusion level was 11% higher than that achieved with the basal diet, whereas feed conversion could be improved by 9% (Table 9). As shown in Figure 4, the response data for the performance criteria weight gain and feed conversion followed a nonlinear trend and thus data were analyzed by simultaneous exponential regression. Regression analysis revealed that responses in weight gain and feed conversion to diluted DLM (65%) were 59 and 47%, respectively, compared with DLM. The 59% bioefficacy value for weight gain fits well with the expected 65% value for this product, whereas the value for feed conversion of 47% cannot be explained based on the current data set or trial conditions. However, this finding demonstrates the extent to which performance data can vary. According to regression analysis,

liquid MHA-FA was only 65 and 49% as effective as DLM regarding weight gain and feed conversion, respectively. Trial 5 Broilers performed extremely well; the maximum responses of weight gain, feed conversion, and breast meat yield (% of carcass) were significantly improved by about 10% to maximum responses of 2,415 g, 1.63, and 28.2%, respectively (Table 10). Responses showed the Met deficiency of the basal diets. Exponential regression revealed the relative effectiveness on an equimolar basis of liquid MHA-FA for weight gain, feed conversion, and breast meat yield from 7 to 40 d of age was 52 (significant), 82 (not significant), and 56% (significant), respectively (Figure 5). In all trials presented here, relative effectiveness values were determined based on a weight/ weight comparison of the 2 Met sources. In trial 5, the supplemented levels of DLM and liquid MHA-FA were adjusted on an equimolar basis [i.e., the corresponding supplemented levels of DLM (treatments 2 to 7) and liquid MHA-FA (treatments 8 to 13) were based on a MHA-FA content of 88% in the commercial product (Table 10)]. To be able to compare all trial data, regression analysis for the response to the different Met sources was conducted on a weight-to-weight basis. Additionally, the data of trial 5 were analyzed based on an equimolar comparison of the

688

JAPR: Research Report

FIGURE 3. Weight gain and feed conversion of broilers fed increasing levels of DL-Met (DLM, 99%), diluted DLM (65%), or liquid methionine hydroxy analog (88%, liquid MHA-FA) from d 1 to 42 (trial 3, Mexican Integrator, Mexico). Values in parentheses indicate 95% confidence intervals. *Feed per gain indicates feed conversion ratio corrected for mortality.

2 products. The results obtained with calculating bioefficacies based on an equimolar comparison for weight gain, feed conversion, and breast meat yield from 7 to 40 d of age were 59 (P < 0.05), 93 (P > 0.05), and 64% (P < 0.05), respectively (data not shown). When multiplied by 0.88, the former values correspond to the same results presented here based on a weight/weight comparison of the 2 Met sources (Figure 5). Therefore, the design of the trial, either equimolar or weight-by-weight comparison of the 2 Met sources, did not affect the estimated relative effectiveness figures. A summary of all 5 broiler trials conducted and presented in this paper is given in Table 11. The studies included a large variety of geo-

graphic locations and climatic areas. The relative effectiveness of liquid MHA-FA compared with DLM was 64% on average, based on the estimated relative effectiveness figures of 11 data sets for weight gain, feed conversion, and breast meat yield. A certain biological variation could be observed in individual experiments, as shown by the relative effectiveness varying from 52 to 82% across all 5 trials, which is common and unavoidable in any biological experiment. The estimated relative effectiveness of the internal standard used to validate the regression analysis (diluted DLM, 65%) relative to DLM for weight gain and feed conversion was 67 and 59%, 69 and 79%, and 59 and 47%, for trials 1, 3, and 4, respectively. On average, these figures

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS

689

TABLE 9. Performance of broiler chickens fed graded levels of DL-Met, diluted DL-Met (65%), or liquid hydroxy analog of Met (MHA-FA) from 7 to 35 d of age (trial 4, University of Arkansas, Fayetteville, AR) Met source

Addition of product (%)

Control DLM DLM DLM DLM DLM Diluted DLM Diluted DLM Diluted DLM Diluted DLM Diluted DLM Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA Liquid MHA-FA

0 0.03 0.06 0.09 0.12 0.15 0.03 0.06 0.09 0.12 0.15 0.03 0.06 0.09 0.12 0.15

Treatment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Weight gain (mean ± SD) 1,579 1,603 1,654 1,710 1,736 1,754 1,602 1,647 1,676 1,683 1,660 1,564 1,619 1,679 1,679 1,716

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

99de 30cde 70abcde 97abc 27a 30a 147de 27abcde 96abcd 90abcd 125abcde 47e 29bcde 196abcd 64abcd 77ab

Feed per gain (mean ± SD) 1.728 1.694 1.656 1.602 1.516 1.583 1.719 1.696 1.646 1.647 1.631 1.710 1.689 1.676 1.647 1.604

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.056a 0.106ab 0.042abcd 0.033cde 0.183e 0.019de 0.064a 0.057ab 0.026abcd 0.036abcd 0.058bcd 0.055ab 0.087abc 0.093abcd 0.045abcd 0.020cd

Means ± SD in a column with no common superscript differ significantly (P < 0.05).

a–e

amount to an estimated relative effectiveness of 63% for diluted DLM (65%) relative to pure DLM. Exponential regression analysis was, hence confirmed to be a statistically valid tool in the present study as suggested by Littell et al. [10] and Lemme et al. [9]. When comparing bioavailability of any essential nutrient, one can only expect to detect differences when the basal diet is clearly deficient in the nutrient to be tested [3, 9]. Additionally, appropriate statistical methods such as simultaneous nonlinear regression analysis [10]

must be used to allow an adequate and unbiased interpretation of the results and estimation of the relative effectiveness of a test substance compared with a reference. Jansman et al. [11] have published an independent literature review on the relative effective of Met sources in poultry and pigs. The result of a relative efficacy of liquid MHA-FA relative to DLM of 64% on average based on 5 broiler trials presented here are in close agreement with the average figure of 68% reported by Jansman et al. [11] for broiler chicks based on 18 trials.

TABLE 10. Performance of broiler chickens fed graded levels of DL-Met or liquid hydroxy analog of Met (MHAFA) from 7 to 40 d of age (trial 5, Federal University of Rio Grande do Sul, Porto Allegro, Brazil) Treatment 1 2 3 4 5 6 7 8 9 10 11 12 13

Met source

Addition of product (%)

Control DLM DLM DLM DLM DLM DLM MHA-FA MHA-FA MHA-FA MHA-FA MHA-FA MHA-FA

0 0.030 0.060 0.100 0.140 0.190 0.240 0.034 0.068 0.114 0.159 0.216 0.273

Weight gain (mean ± SD) 2,183 2,276 2,349 2,396 2,412 2,412 2,415 2,245 2,286 2,369 2,380 2,390 2,385

± ± ± ± ± ± ± ± ± ± ± ± ±

12d 40bcd 39abc 62ab 36a 48a 64a 52cd 59abcd 34abc 33ab 59ab 61abc

Feed per gain (mean ± SD) 1.766 1.724 1.700 1.679 1.636 1.631 1.634 1.736 1.693 1.668 1.667 1.635 1.627

± ± ± ± ± ± ± ± ± ± ± ± ±

0.016a 0.029ab 0.029abc 0.027bc 0.043c 0.029c 0.022c 0.013ab 0.003abc 0.009bc 0.021bc 0.044c 0.038c

Means ± SD in a column with no common superscript differ significantly (P < 0.05).

a–d

Breast meat yield (% of carcass; mean ± SD) 25.8 26.9 27.0 27.4 27.6 28.0 28.2 26.0 26.7 26.8 27.4 27.8 28.1

± ± ± ± ± ± ± ± ± ± ± ± ±

0.7b 1.0ab 0.7ab 0.8ab 0.7ab 0.4a 0.2a 0.1b 0.8ab 1.3ab 0.5ab 0.7ab 0.5a

690

JAPR: Research Report

FIGURE 4. Weight gain and feed conversion of broilers fed increasing levels of DL-Methionine (DLM, 99%), diluted DLM (65%), or liquid methionine hydroxy analog (88%, liquid MHA-FA) from d 7 to 35 (Trial 4, University of Arkansas, USA). Values in parentheses indicate 95% confidence intervals. *Significantly less than 88%; P < 0.05.

The physiological reasons for the lower relative efficacy of liquid MHA-FA have also been investigated. Differences in relative bioefficacy may be explained by reduced intestinal absorption of MHA-FA, inefficient conversion of MHA-FA to Met after absorption or a combination of both of these factors. The differences in bioefficacy due to intestinal absorption have been studied using several methods. Han et al. [21] determined that absorption of DLM and DL-MHA-FA is equal using a true-digestibilitybalance assay and cecectomized broiler chickens. This method measures the nonabsorbed MHA-FA and DLM present in digesta after in-

testinal passage. However, this methodology is questionable because metabolism of these compounds to other products by intestinal bacteria would result in an overestimation of the absorption of both compounds. Several studies with broilers using radiolabeled Met sources indicated a significantly lower absorption of MHAFA compared with DLM [22, 23]. In agreement with Esteve-Garcia and Austic [1], Maenz and Engele-Schaan [23] concluded that there is a substantial conversion of dietary MHA-FA during passage through the small intestine to compounds that cannot be used as a source of Met by the bird. This might be due to degradation of

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS

691

FIGURE 5. Weight gain, feed conversion, and breast meat yield of broilers fed increasing levels of DL-Met (DLM, 99%) or liquid methionine hydroxy analog (88%, liquid MHA-FA) from d 7 to 40 (trial 5, Federal University of Rio Grande do Sul, Porto Allegro, Brazil). Values in parentheses indicate 95% confidence intervals. *Significantly less than 88%; P < 0.05.

a substantial fraction of MHA-FA by microbial fermentation during passage through the small intestine. Moreover, oligomers of liquid DL-

MHA-FA are poorly absorbed. About 23% of the MHA-FA molecule fraction is present in the form of dimers and oligomers, which have a

JAPR: Research Report

692

TABLE 11. Estimated effectiveness of liquid hydroxy analog of Met (MHA-FA) relative to based on weight gain, feed conversion, and breast meat yield of broiler chickens Trial and location 1, Research Centre Foulum, Tjele, Denmark 2, Perdue Farms Inc., Salisbury, MD 3, Mexican Integrator, Mexico 4, University of Arkansas, Fayetteville, AR 5, Federal University of Rio Grande do Sul, Porto Allegro, Brazil Overall mean, n = 11 1

in trials 1 to 5

Weight gain

Feed conversion

Breast meat yield

641 70 63 65 521

671 66 73 491 82

— — — — 561

Relative effectiveness of liquid MHA-FA was significantly lower than that of

reduced nutritional value. This fact has been known for a long time and is noted in the original MHA-FA patent [24]. Van Weerden et al. [7] reported significantly lower relative effectiveness of DL-MHA-oligomers compared with the commercial product mix of mono-, di-, and oligomers, which in turn was 66% as efficacious as DLM. Furthermore, both products are absorbed by specific active absorption mechanisms, whereas absorption by simple diffusion is negligible [22, 25]. There is evidence that the mechanism responsible for uptake of the DLMHA molecules is less efficient than that for DLM uptake [25]. In addition to the poor nutritive value of MHA-FA oligomers, DLM was found to be faster and more efficiently absorbed than MHA-FA monomers. Three previous studies with broiler chickens comparing the efficiency of absorption of MHA-FA and DLM have reported that between 10 to 20% of the original radiolabeled MHA-FA activity in feed is found in the distal sections of the small intestine compared with 4 to 5% for DLM [1, 22, 25]. Very recently, Drew et al. [26] investigated the absorption of 3H-labeled L-Met and L-2-hydroxy4-methylthiobutanoic acid (L-MHA-FA) in germ-free and conventional broiler chickens to determine the effect of intestinal bacteria on the absorption of Met and MHA-FA. The authors found that more than 10% of the 3H-MHA activity remained in the distal ileum of the conventional birds compared with 3.7% of 3H-Met activity at the same location. In contrast, the residual 3H-MHA activity in broiler chickens raised completely germ-free was significantly decreased compared with conventional birds. This finding confirms the initial hypothesis by Drew et al. [26] that absorption and metabolism of 3 H-MHA by intestinal bacteria has a significant

DL-Met

64 DL-Met

(see Figures 1 to 5 for details).

effect on the decreased intestinal absorption of MHA-FA relative to DLM. Uptake of DLM and MHA-FA across the brush border membrane takes place by 2 different transport systems [23]. Met is transported by the system B amino acid transporter, whereas MHA-FA absorption is transported via a +H-dependent non-stereospecific transport system. The same study also reported that L-Met had a higher affinity for its transporter and a higher maximal velocity of transport than L-MHA. This result suggests that Met is removed more quickly from the intestinal lumen than MHA-FA and therefore has less exposure to intestinal bacterial resulting in decreased bacterial uptake or degradation. Based on these results, it is reasonable to assume that there are differences in intestinal transport of MHA-FA and Met that may increase the availability of MHA-FA to gut bacteria. Because only L-Met is used in protein metabolism, D-MET, D-MHA-FA, and L-MHA-FA must be converted to L-Met [27]. This conversion takes place in 2 steps. First, the α-carbon of all 3 compounds is oxidized to yield ketoMet [28], and the keto-Met is then converted to L-Met. The enzymes responsible for these reactions are found primarily in the liver and kidney [28], and the activity of these enzymes does not appear to be limiting for growth in broiler chickens. Dibner and Ivey [27] showed a 7-fold excess of enzyme activity in the liver alone for the conversion of D-MET, D-MHA-FA, and L-MHA-FA to keto-Met. However, because these reactions require energy for the conversion of these compounds to keto-Met, DL-Met has the advantage of requiring a conversion of only 50% of the ingredient (D-Met), compared with 100% of DL-MHA-FA. The relative contribution of the energy requirements for these conversions to the

HOEHLER ET AL.: METHIONINE SOURCES IN BROILER CHICKENS

693

difference in bioefficacy of DL-Met relative to DL-MHA-FA has not yet been determined.

CONCLUSIONS AND APPLICATIONS 1. The results of the presented 5 broiler experiments suggest that simultaneous exponential regression analysis is a valid statistical means for determination of relative bioefficacy of Met sources. 2. The relative effectiveness of liquid MHA-FA is significantly lower than that of DLM in broiler chickens. The average bioefficacy value estimated in this series of diverse experiments conducted in different parts of the world was 64% for liquid MHA-FA on average on a weight/weight basis of the 2 Met sources. 3. This finding agrees very well with the results of an independent literature study, which reported that the average bioefficacy of liquid MHA-FA compared with DLM was 68% on a weight/ weight basis or 77% on an equimolar basis of the 2 Met sources in broiler chickens.

REFERENCES AND NOTES 1. Esteve-Garcia, E., and R. E. Austic. 1993. Intestinal absorption and renal excretion of dietary methionine sources by the growing chicken. J. Nutr. Biochem. 4:576–587. 2. Esteve-Garcia, E., and L. Llaurado. 1997. Performance, breast meat yield and abdominal fat deposition of male broiler chickens fed diets supplemented with DL-methionine or DL-methionine hydroxy analog free acid. Br. Poult. Sci. 38:397–404. 3. Huyghebaert, G. 1993. Comparison of DL-methionine and methionine hydroxy analog-free acid in broilers by using multiexponential regression models. Br. Poult. Sci. 34:351–359. 4. Rostagno, H. S., and W. A. Barbosa. 1995. Biological efficacy and absorption of DL-methionine hydroxy analog free acid compared to DL-methionine in chickens as affected by heat stress. Br. Poult. Sci. 36:303–312. 5. Schutte, J. B., and J. de Jong. 1996. Biological efficacy of DL-methionine hydroxy analog-free acid compared to DL-methionine in broiler chicks as determined by performance and breast meat yield. Agribiol. Res. 49:74–82.

13. National Research Council. 1994. Nutrient Requirements of Poultry. 9th rev. ed. Natl. Acad. Press, Washington, DC. 14. Perdue 5632 broilers, Perdue Farms, Inc., Salisbury, MD. 15. Hybro broilers, Euribred, The Netherlands, Celaya, Guanajuato, Mexico. 16. Ross 308 broilers, Ross Breeders, Sallisaw, OK. 17. Ross 308 broilers, Cooperative Languiru, Teutonia, Rio Grande do Sul, Brazil. 18. Llames, C. R., and J. Fontaine. 1994. Determination of amino acids in feeds: Collaborative study. J. AOAC Int. 77:1362–1402. 19. Naumann, C., R. Bassler, R. Seibold, and C. Barth. 1997. Methodenbuch Band III. VDLUFA-Verlag, Darmstadt, Germany. 20. SAS Institute. 2000. SAS/STAT User’s Guide. Version 8 ed. SAS Inst. Inc., Cary, NC. 21. Han, Y., F. Castanon, C. M. Parsons, and D. H. Baker. 1990. Absorption and bioavailability of DL-methionine hydroxy analogue compared to DL-methionine. Poult. Sci. 69:281–287.

6. Thomas, O. P., C. Tamplin, S. D. Crissey, E. Bossard, and A. Zuckerman. 1991. An evaluation of methionine hydroxy analog free acid using a nonlinear (exponential) bioassay. Poult. Sci. 70:605–610.

22. Lingens, G., and S. Molnar. 1996. Studies on metabolism of broilers by using 14C-labelled DL-methionine and DL-methionine hydroxy analogue Ca-salt. Arch. Anim. Nutr. 49-113–124.

7. Van Weerden, E. J., J. B. Schutte, and H. L. Bertram. 1992. Utilization of the polymers of methionine hydroxy analogue free acid (MHA-FA) in broiler chicks. Arch. Geflgelkd. 56:63–68.

23. Maenz, D. D., and C. M. Engele-Schaan. 1996b. Methionine and 2-hydroxy-4-methylthiobutanoic acid are partially converted to non-absorbed compounds during passage through the small intestine and heat exposure does not affect small intestinal absorption of methionine sources in broiler chicks. J. Nutr. 126:1438–1444.

8. Wallis, I. R. 1999. Dietary supplements of methionine increase breast meat yield and decrease abdominal fat in growing broiler chickens. Aust. J. Exp. Agric. 39:131–141. 9. Lemme, A., D. Hoehler, J. J. Brennan, and P. F. Mannion. 2002. Relative Effectiveness of Methionine Hydroxy Analog Compared to DL-Methionine in Broiler Chickens. Poult. Sci. 81:838–845. 10. Littell, R. C., P. R. Henry, A. J. Lewis, and C. B. Ammermann. 1997. Estimation of relative bioavailability of nutrients using SAS procedures. J. Anim. Sci. 75:2672–2683. 11. Jansman, A. J. M., C. A. Kan, and J. Wiebenga. 2003. Comparison of the biological efficacy of DL-methionine and hydroxy-4methylthiobutanoic acid (HMB) in pigs and poultry. Page 55 in Documentation Rep. no. 29. Centraal Veevoederbureau (CVB, Central Bureau for Livestock Feeding), Lelystad, The Netherlands. 12. Ross 208 broilers, Faellesrugeriet a.m.b.a., Randers, Denmark.

24. Monsanto Company. 1955. Preparation of methionine analogues. US Patent 3,272,860. 25. Maenz, D. D., and C. M. Engele-Schaan. 1996a. Methionine and 2-hydroxy-4-methylthiobutanoic acid are transported by distinct + Na -dependent and H +-dependent systems in the brush border membrane of the chick intestinal epithelium. J. Nutr. 126:529–536. 26. Drew, M. D., A. G. Van Kessel, and D. D. Maenz. 2003. Absorption of methionine and 2-hydroxy-4-methylthiobutanoic acid in conventional and germ-free chickens. Poult. Sci. 82:1149–1153. 27. Dibner, J. J., and F. J. Ivey. 1992. Capacity of the liver of the broiler chick for conversion of supplemental methionine activity to L-methionine. Poult. Sci. 71:700–708. 28. Dibner, J. J., and C. D. Knight. 1984. Conversion of 2hydroxy-4-(methylthio)butanoic acid (HMB) to L-methionine in the chick: A stereo-specific pathway. J. Nutr. 114:1716–1723.