The 15N-leucine single-injection method allows for determining endogenous losses and true digestibility of amino acids in cecectomized roosters

This study was conducted to assess the influence of dietary protein content in poultry when using the 15N-leucine single-injection method to determine endogenous amino acid losses (EAALs) in poultry. Forty-eight cecectomized roosters (2.39 ± 0.23 kg) were randomly allocated to eight dietary treatments containing protein levels of 0, 3%, 6%, 9%, 12%, 15%, 18% and 21%. Each bird was precisely fed an experimental diet of 25 g/kg of body weight. After feeding, all roosters were subcutaneously injected with a 15N-leucine solution at a dose of 20 mg/kg of body weight. Blood was sampled 23 h after the injection, and excreta samples were continuously collected during the course of the 48-h experiment. The ratio of 15N-enrichment of leucine in crude mucin to free leucine in plasma ranged from 0.664 to 0.763 and remained relatively consistent (P > 0.05) across all treatments. The amino acid (AA) profiles of total endogenous AAs, except isoleucine, alanine, aspartic acid, cysteine, proline and serine, were not influenced (P > 0.05) by dietary protein contents. The predominant endogenous AAs in the excreta were glutamic acid, aspartic acid, threonine, serine and proline. The order of the relative proportions of these predominant AAs also remained relatively constant (P > 0.05). The endogenous losses of total AAs determined with the 15N-leucine single-injection method increased curvilinearly with the dietary protein contents. The true digestibility of most AAs and total AAs was independent of their respective dietary protein levels. Collectively, the 15N-leucine single-injection method is appropriate for determining EAALs and the true digestibility of AAs in poultry fed varying levels of protein-containing ingredients.


Introduction
The accurate quantification of endogenous amino acids losses (EAALs) in the intestines of animals is crucial for determining amino acid (AA) requirements and for calculating the true AA digestibility of feedstuffs. In poultry, EAALs have traditionally been determined by the fasted PLOS  performed as described by Payne et al. [16]. All birds were individually raised in a single cage under a 16 h light and 8 h dark photoperiod. Drinking water was offered ad libitum. The birds were randomly divided among 8 dietary treatments. The experimental diets included a nitrogen-free diet and seven protein-containing diets (Table 1). Soybean meal was the only source of dietary CP. The AA composition of the diets is shown in S1 Table. All other nutrients and energy met or exceeded the estimated requirements for roosters [17]. A fiber source was used to equalize the crude fiber levels in all experimental diets. The electrolyte balance was constant across all the diets recommended by Adedokun et al. [1]. 15

N-Leucine injection and sampling
The general outline of the 15 N-leucine injection and sampling protocol is summarized in S2 Table. The experiment included three periods. In the first period, the birds were adapted to the experimental diets for 5 consecutive days. On day 6, the roosters were deprived of the diet for 24 h to empty the digestive tract of all dietary residues. Then, the birds were precision-fed 25.0 g of dry matter/kg of body weight of each experimental diet using the method presented by Kim    [18]. Following the precision feeding, excreta were collected quantitatively for 48 h. A bag (250 mL) was attached to a hollowed-out plastic bottle cap fitted to the rooster's cloaca. Then, 10 mL of a 10% formic acid solution was added to the bags to inhibit enzyme and microbial activity in the excreta. The excreta in the bag was collected once every two hours, and frozen (−40˚C) immediately. The 48-h excreta samples were pooled for each bird when the experiment was over. Blood samples (5 mL) were sampled 23 h after precision feeding [3]. The above blood and excreta samples were used to determine the basal 15 N-enrichment of leucine in the birds. In the second period, all birds were given one week to recover from the first period. In the third period, all roosters were subcutaneously injected with 15 N-L-leucine (98% 15 N-enrichment; Cambridge Isotope Labs. Inc., Andover, MA, USA) at a dose of 20 mg/kg of body weight after precision feeding. The other procedures were the same as those described for the first period.

Sample preparation
The sample preparation protocol was adapted from the study by Steendam et al. [19]. Briefly, blood samples were centrifuged (10 min, 2000×g, 4˚C) to recover plasma and then treated twice with a 10% (W/V) trichloroacetic acid solution to obtain deproteinized plasma. The AAs in the deproteinized plasma were purified though passage on AG 50W-X8 cation-exchange resin (hydrogen form, 200 μm, Sigma, St. Louis, MO, USA). The AAs were eluted with 4 mol/ L fresh NH 4 OH. After NH 4 OH evaporation, the samples were re-dissolved with 5 mL of ultrapure water for the 15 N analysis. The excreta samples were thawed, homogenized, freeze-dried, weighed and then carefully ground to pass through a 1 mm screen for further analysis. Crude mucin in the excreta was isolated with the ethanol precipitation method described by Lien et al. [5] and Leterme et al. [20]. Hydrolyzed excreta or crude mucin samples that were obtained using hydrochloric acid method (10 mL of 6 mol/L HCl, 110˚C, 24 h) were purified similarly to the plasma samples for 15 N analysis.

Chemical and isotope analyses
Dry matter, CP (N × 6.25) and neutral detergent fiber were analyzed in the ingredients and diets [21]. The AA concentrations of the ingredients, diets and excreta samples were analyzed using the Biochrom 30 amino acid analyzer (Pharmacia Biotech, Cambridge, UK). The 15 Nenrichment of leucine was determined with a gas chromatograph coupled with a combustion oven and an isotope ratio mass spectrometer (GC-C-IRMS, Thermo-Finngan, USA) system according to a previously presented study [4]. The AAs were derivatized with thionyl chloride and pivaloyl chloride [22], which led to the formation of N-pivaloyl-isopropyl derivatives. The samples (1.0 μL, split mode = 10:1) were injected onto an HP ULTRA 2 column (50 m × 0.32 mm × 0.52 μm) at an injector temperature of 250˚C. The column oven was initially set at 70˚C (for 1 min), rise at a rate of 3˚C/min to 220˚C, and then 10˚C/min to 300˚C with a holding time of 8 min. The temperatures of the CuO/Pt combustion reactor and reduction oven were 850˚C and 650˚C, respectively.

15
N-enrichment in leucine. The 15 N-enrichment (atom percent excess, APE) in leucine was obtained according to formulas (1) and (2), which were adapted from Wolfe et al. [23]: TTRð%Þ ¼ ½ð15N=14NÞ sam À ð15N=14NÞ bas Â ð1 À AÞ ð2Þ where the 15N/14N ratios originate from the ratios of the m/z 29 to m/z 28 ion current signals in a labeled sample (sam) or in basal sample (bas) [22]. The value of A is the natural abundance of 15 N. The 15 N-leucine single-injection method. The calculation principle of this method, which has been shown previously [3], is expressed as the formula (3): where i denotes the number of experimental protein-containing diets, and NL e0 /NL p0 and NL ei /NL pi are the 15 N-enrichment ratios of endogenous leucine in the excreta and free leucine in the deproteinized plasma from the birds fed either nitrogen-free diet or the i th protein-containing diet, respectively. The 15 N-leucine content in the excreta was derived from endogenous 15 N-leucine; therefore, we obtained formula (4). By combining formulas (3) and (4), we generated the formula (5): where NL ex represents 15 N-enrichment of total leucine in the excreta; W e and W t (mg/kg of dry matter intake (DMI)) are the losses of endogenous leucine and total leucine in the excreta, respectively. The AA profile of total endogenous AAs was assumed to be relatively stable [4,13]. Therefore, the AA profile obtained from the nitrogen-free diet could be used to calculate the endogenous losses of other AAs in birds fed protein-containing diets.
The gradient protein method for estimating the total endogenous AAs profile. Apparent digestibility (AD) and true digestibility (TD) of AAs in diets were calculated according to formulas (6) and (7): where TI, TF and E denote the total AA input (mg/kg of DMI), total fecal AA output (mg/kg of DMI) and EAALs (mg/kg of DMI) from the experimental diets, respectively. When the source of dietary protein and the contents of the other dietary components, including fiber and anti-nutritive factors, are similar among the experimental diets, then the TD will be the same at varying dietary CP levels [2,15]. Therefore, Based on several previous studies [24][25][26], we assumed that the dietary CP levels, which varied within a small range (3%), had little effect on the EAALs.
According to formulas (6), (7), (8) and (9), we obtained formula (10): where E iðiþ1Þ (mg/kg of DMI) represents the mean endogenous losses of AAs between the i th and (i+1) th protein-containing diets. The profile of the total endogenous AAs, expressed as a proportion of total AAs, was calculated as follows: where P iðiþ1Þ and TE iðiþ1Þ (mg/kg of DMI) represented mean proportion of each individual AA and the total endogenous losses of AAs between the i th and (i+1) th protein-containing diets, respectively.

Statistical analysis
Statistical analysis was performed using SPSS 21.0 (IBM-SPSS Inc., Chicago, Il, USA). The data are presented as the means and pooled standard error (SEM) (n = 6). Differences between the means of all groups were compared by one-way ANOVA, followed by Bonferroni corrections test. Significant differences were declared at P<0.05.

Results
The values of 15 N-enrichment of leucine in the deproteinized plasma, crude mucin and excreta are presented in Table 2. The values of 15 N-enrichment were significantly different in the different samples for all treatments. The highest 15 N-enrichment was observed in the deproteinized plasma, followed by the crude mucin and then the excreta. All the values of 15 N-enrichment significantly decreased (P < 0.05) when the dietary CP level was increased to 15%. The ratios of 15 N-enrichment in the different samples are also shown in Table 2. The value for NL cm /NL p ratios ranged from 0.664 to 0.763, whereas the NL ex /NL p ratios ranged from 0.262 to 0.744. The dietary CP levels had no influence (P > 0.05) on the NL cm /NL p ratios. The contribution of endogenous leucine to total leucine and the endogenous losses of leucine in the excreta are presented in Table 3. No significant differences (P > 0.05) were found between contributions calculated with NL e0 /NL p0 and NL cm /NL p . Endogenous leucine losses calculated with NL e0 /NL p0 increased significantly (P < 0.05) when the dietary CP level increased to 12%.
The AA profile of the total endogenous AAs (expressed as a proportion of total endogenous AAs) determined with the gradient protein method is shown in Table 4. The ratios of most of the indispensable AAs (except isoleucine) and glutamic acid were not influenced (P > 0.05) by the dietary CP levels. The proportions of isoleucine, aspartic acid and serine in the total endogenous AAs were greater (P < 0.05) in diets with increasing CP levels from 12% to 21% than those in the nitrogen-free diet, whereas the proportions of alanine, cysteine and proline were lower (P < 0.05). No significant differences (P > 0.05) were observed in the ratios of most AAs in the total endogenous AAs when the dietary CP levels were increased from 3% to 12% or from 12% to 21%. The order of the relative proportions of these predominant AAs was similar across all dietary treatments.
The endogenous losses of other AAs calculated with the AA profile of total endogenous AAs in the nitrogen-free diet are presented in Table 5. The dietary CP levels had a significant    https://doi.org/10.1371/journal.pone.0188525.t003 effect (P < 0.05) on the endogenous loss of most AAs (except histidine). The endogenous loss of total AAs in roosters fed the 21% CP diet was approximately 2.2 times higher than the endogenous losses determined with birds fed the nitrogen-free diet. Similar endogenous losses (P > 0.05) of all individual AAs were observed when increasing the dietary CP levels from 0 to 6%. The endogenous losses of most individual AAs and the total AAs increased slightly (P > 0.05) when the dietary CP levels were increased from 0 to 6% or from 12% to 21%. The endogenous losses of most individual AAs and the total AAs increased dramatically when the dietary CP levels were increased from 6% to 12% (especially from 6% to 9%).
The true AA digestibility calculated with the EAALs determined by the 15 N-leucine singleinjection method (total losses) was relatively constant except for a few AAs that had lower digestibility at dietary CP levels of 18% or 21% ( Table 6). The apparent digestibility of the total AAs increased nonlinearly with the dietary CP levels, whereas the true digestibility of total AAs was independent of the dietary CP levels (Fig 1).

Discussion
One of the aims of the present study was to evaluate the 15 N-leucine single-injection method within a wide range of dietary CP levels. The results showed that the NL cm /NL p ratios were not influenced by dietary CP levels. Meanwhile, the contributions of endogenous leucine to total leucine in the excreta calculated as NL e0 /NL p0 and NL cm /NL p were similar at the same dietary CP level. Despite some changes in individual AA proportions (mainly dispensable AAs) when Table 4. Amino acid profile (%) of total endogenous amino acids calculated with the gradient protein method in precision-fed cecectomized roosters fed the nitrogen-free diet and soybean meal diets at varying crude protein ranges 1 . the dietary CP levels were increased from 0 to 21%, the AA profile of total endogenous AAs was relatively constant in the present study. These findings support our assumptions. The tracer can be given as either a single bolus or continuously when using the isotope labeling technique to trace endogenous AAs [23]. Continuous 15 N-leucine infusion is the most commonly used technique to determine EAALs in pigs and rats [4][5][6]. This method assumes that the 15 N-enrichment of endogenous leucine in excreta is approximately equal to that in plasma (i.e., NL e = NL p ) after continuous 15 N-infusion for 7 to 8 d. Instead of using continuous 15 N-leucine infusion to obtain a steady state, we proposed the NL cm /NL p ratio remained constant when testing another way of using a single bolus. By regression analysis of the NL p value and time in our previous study, we found the optimum time of blood sampling in our method was 23 h after the isotope injection [3]. A single blood sample taken at this time could be used to reflect the weighted mean value of definite integral NL p during the sampling period. Additionally, excreta samples from one bird were pooled to obtained the NL cm value, which was more representative than the 15 N-enrichment at a single time point. Therefore, the 15 N-leucine injection and sample preparation in our method are more economical and simpler than the 15 N-isotope infusion method.

P-Value
After a single injection, 15 N-leucine becomes a part of the pool of the total blood leucine and has the same metabolic rate as unlabeled leucine [3]. Thus, the ratio of labeled to unlabeled endogenous leucine in the excreta has the same tendency to vary as the ratio of labeled to Table 5 unlabeled leucine in the different sources of endogenous AAs (e.g., plasma, mucosa, or desquamated cells). In other words, the NL cm /NL p ratio indicates the NL e /NL p ratio, which was also confirmed by our present data. Moreover, we proposed a gradient protein method to obtain the AA profiles of total endogenous AAs in birds fed protein-containing diets. The gradient protein method depends on the theory that true AA digestibility is independent of the dietary CP level [2,14]. Based on several previous studies [24][25][26], we assumed that variation in the dietary CP levels within a small range had little effect on the EAALs. A study in pigs fed diets with equal graded protein levels of 4% found no significant differences in the EAALs among the adjacent CP levels. In the present study, the EAALs were assumed to be relatively constant or to vary little when the dietary CP levels varied within 3%. Many studies have generally assumed that the AA profiles of endogenous protein are constant [4,6,27]. No significant differences were observed in the ratios of most AAs in the endogenous protein in broiler chickens when increasing the dietary enzyme-hydrolyzed casein concentration from 5% to 20% [24]. This result was almost in agreement with the present data in our study. Although increasing the dietary CP levels stimulates the secretion of digestive enzymes, sloughed cells and mucin [24], the degree of this stimulation source is similar for each secretion. Therefore, the proportion of each excretion source in the total endogenous excretion will not change. Meanwhile, although the relative contribution of each source may vary, the AA composition of a single source tends to be constant [28]. Thus, we infer that the AA profiles of total endogenous AAs are not influenced by dietary CP levels. Moreover, glutamic acid, aspartic acid, threonine, proline and serine dominated the AA profiles of total  [5,27]. Additionally, a summary by Boisen et al. [31] indicated that the AA profiles of endogenous protein, which were obtained from different diets and different methods of determination, remained relatively stable. Therefore, the AA profiles of total endogenous AAs in birds fed a nitrogen-free diet were generally acceptable for use in the calculations. Notably, endogenous losses of alanine, cysteine and proline may be underestimated at high dietary CP levels due to the lower proportions of these AAs when the AA profiles of total endogenous AA in the nitrogen-free diet treatment are used. Another purpose of this study was to estimate total EAALs using the 15 N-leucine singleinjection method and to obtain the true digestibility of AAs in cecectomized roosters fed soybean meal diets with varying CP levels. The results showed that endogenous losses of total AAs increased in response to increases in dietary CP levels, and reached relative equilibrium above 12% dietary CP. Similar changes were found in previous studies [24,32]. The dietary protein or AA content is one of the main factors affecting EAALs [1]. An increasing dietary CP content likely results in higher specific endogenous losses of AAs, which may be responsible for this result. Moreover, although the apparent AA digestibility nonlinearly increased with the dietary CP level, the true AA digestibility was independent of the respective dietary CP. This finding is in accordance with a well-known conclusion that has been widely demonstrated [2,15]. This finding also indirectly proves that the 15 N-leucine single-injection method can be an effective means for determining total EAALs in poultry fed varying levels of protein-containing ingredients.

Conclusions
The present data support the assumptions of the 15 N-leucine single-injection method within a wide range of dietary CP levels. The results show that this method allows for the determination the total EAALs and true AAs digestibility in poultry fed protein-containing diets. Notably, endogenous losses of alanine, cysteine and proline may be underestimated at high dietary CP levels. Further studies are also necessary to investigate the accuracy and repeatability of this method, especially using other ingredients as the source of dietary protein.
Supporting information S1 Table. Analyzed amino acid compositions (g/kg as-fed basis) of the experimental diets. (DOCX) S2 Table. Experimental scheme for the 15 N-leucine single-injection method, including caecectomy surgery, injection of 15 N-L-leucine, and sampling times of blood and excreta. (DOCX)