Comparison of two in vitro methods progressed in a computer-controlled simulated digestion system to determine amino acid digestibility of feed ingredients for yellow-feathered roosters

This experiment compared amino acid (AA) digestibility assessed by 2 in vitro methods using a computer-controlled simulated digestion system and in vivo assay for corn, soybean meal, casein, corn gluten meal, cottonseed meal, rapeseed meal and a corn-soybean meal diet. In vitro method 1 simulated gizzard digestion at pH 2.0, followed by small intestinal digestion, and the subsequent clearance of the digested product from dialysis tubing. In vitro method 2 was similar to the first method, except that pH in gizzard digestion was 3.5 and there was an enzymatic inactivation stage before digested product clearance. Each in vitro method included 5 replicates per treatment, with 1 digestion tube per replicate. Cecectomized Chinese yellow-feathered roosters (average body weight of 2.73 kg) were assigned to 1 of the following treatments: corn, soybean meal, or a corn-soybean meal diet (n = 6 replicates of 3 roosters per treatment); or casein, corn gluten meal, cottonseed meal, or rapeseed meal (n = 5 replicates of 3 roosters per treatment) in a completely randomized design to assess digestibility of AA. The relative deviation was within 5% comparing methods 1, 2 and in vivo method for 87.5% and 92.0% of all AA digestibility measured of 7 samples, respectively. Significant linear relationships were observed between in vitro methods 1 and 2 for the digestibility of 14 AAs (except for Cys) and total amino acid (TAA) (r ≥ 0.778; P < 0.05). Significant linear relationships were found between in vitro method 1 and in vivo results for 9 AAs (except for His, Ile, Asp, Cys, Glu and Ser) and TAA (r ≥ 0.866; P < 0.05). Similarly, significant linear relationships between in vitro method 2 and in vivo findings were observed for 11 AAs (except for His, Lys, Cys and Glu) and TAA (r ≥ 0.776; P < 0.05). The linear regression of in vivo assay on in vitro method 1 or 2 overlapped with Y = X for 7 AA and TAA or 10 AA and TAA, respectively. Our findings suggest in vitro method 2 is superior to method 1 for estimating AA digestibility of yellow-feathered roosters, this indicates that gizzard pH or inactivation of enzymes modulates the effectiveness of in vitro digestibility assays.