Determination of standardized ileal digestibility of crude protein and amino acids and digestible indispensable amino acid score of faba beans, lentils, and yellow peas fed to growing pigs

Abstract The objective of this study was to determine the standardized ileal digestibility (SID) of protein and amino acids (AA) and calculate the digestible indispensable AA score (DIAAS) of three varieties of Canadian grown pulses (faba bean, lentil, and yellow pea). Three steam-pelleted (80 °C) diets (faba/lentil/pea; 40% inclusion) and a nitrogen-free diet were fed to eight cannulated growing pigs in a 4 × 4 replicated Latin square design. The SID values were used to calculate DIAAS. There were no differences in SID of protein and AA when all diets were compared (P > 0.05). The SID of methionine was 95% for faba beans, 96% for lentils, and 96% for peas. The SID of lysine was 93% for faba beans, 92% for lentils, and 90% for peas. Tryptophan had the lowest SID across all indispensable AA of all pulses (59% faba, 61% lentil, and 41% pea). All other indispensable AA SID were above 80%. Tryptophan was the limiting indispensable AA for all pulses as determined by DIAAS, with scores of 30 (faba), 57.83 (lentil), and 27.27 (pea). These values will aid in optimal inclusion of steam-pelleted pulses in diets for swine and provide an insight on the protein quality of pulses for humans.


Introduction
There is growing interest in the use of alternative/novel ingredients, such as pulse crops (peas, beans, lentils, chickpeas, etc.), as protein sources in livestock feed.However, while pulses are nutrient rich, their use in swine feed is minimal compared with other common feed ingredients, such as soybean meal, due to a dearth of data on their nutrients and characteristics when incorporated into pig diets.
Pulses are defined as legumes harvested for the dried seed (FAO 2011a).It is primarily the contribution of energy and protein that make pulses an attractive ingredient as animal feed (Sherasia et al. 2017).However, plant proteins, including pulses, tend to have lower content of indispensable amino acids (AA), including sulfur AA (SAA) (methionine [Met] and cysteine [Cys]) and tryptophan (Trp; Young and Pellet 1994;Singh 2017), and may not be able to supply adequate levels of indispensable AA required by the animal.Additionally, due to the presence of anti-nutritional factors (ANF) in pulses, these ingredients should be thermal processed prior to feeding (e.g., pelleting or extrusion) to decrease ANF and enhance digestibility.While there has been some research conducted on the digestibility of pulse ingredients fed to pigs (Mariscal-Landín et al. 2002;Friesen et al. 2006;Stein and Bohlke 2007;Jezierny et al. 2010;Presto et al. 2011;Eklund et al. 2012;Nørgaard et al. 2012;Han et al. 2020;Hugman et al. 2021aHugman et al. , 2021b)), pulses have not been extensively researched in swine diets, especially considering the diversity of pulse ingredients.Moreover, plant breeding continues to advance the nutrient composition of feed stuffs, such as zero tannin pulse varieties of faba beans.Thus, digestibility values for pulses will be required to be updated as feed ingredients change and evolve.
The use of pulses as a source of dietary protein is not limited to animal feed, as they are also an important food crop for human consumption.To guide human consumers on quality sources of protein, protein claims are added to nutritional labels.In 2011, The FAO recommended that the digestible indispensable AA score (DIAAS) be used to determine the protein quality of human foods via scores (FAO 2011b).The DIAAS uses the ileal digestibility values for each indispensable AA to determine the lowest scoring indispensable AA and is defined as: DIAAS (%) = 100 × [(mg of digestible dietary indispensable AA in 1 g of the dietary protein)/(mg of the same dietary indispensable AA in 1 g of the reference protein)].Based on the lowest scoring AA value, a protein claim can then be made for the ingredient of interest if a score is ≥75 to guide human selection and consumption of quality protein sources.Given the difficulty in sampling ileal digesta from humans, the growing pig was identified as a suitable model for generating ileal digestibility values for human nutrition (FAO 2011b).Consequently, swine are considered an important and appropriate animal model for providing an insight on protein quality of ingredients for human nutrition.
There is limited literature available for ileal digestibility of crude protein and AA for an assortment of pulses and their respective varieties.Further delineation of this data could be beneficial across swine and human diets as food systems transition to include more sources of plant protein.Therefore, the objective of this study was to determine the SID values of protein and AA of steam-pelleted faba beans, lentils, and yellow peas pelleted in growing pig diets and to determine the DIAAS of each steam-pelleted pulse.

Materials and methods
The experimental protocol and study design were reviewed and approved by the University of Guelph Animal Care Committee (AUP# 4439).Handling and caring of the animals were in accordance with the Canadian Council on Animal Care guidelines (CCAC 2009).An unpublished version of this article can be accessed via a thesis (Cargo-Froom 2022).

Animals, housing, and ileal cannulation
Eight healthy Yorkshire barrows (BW 18.32 ± 0.78 kg) were obtained from Arkell Swine Research Station (Guelph, ON, Canada).Pigs were pair-housed for the first 7 days to allow for environmental adaptation.Pigs were housed in plexi-glass pens with tenderfoot flooring, in a temperature-controlled room (23-24 • C) with 12 h light and 12 h dark cycles, at the Department of Animal Biosciences, University of Guelph (Guelph, ON, Canada).Pens were cleaned once daily throughout the trial.A commercial feed (FFM#1 Pig Starter Ration, Floradale Feed Mill Limited, Floradale, ON, Canada) and water were provided ad libitum.Environmental enrichment (e.g., toys and chains) were provided throughout the entire study.Prior to ileal cannulation surgery, feed was removed and pigs were fasted overnight (min 14 h).A simple T-cannula was surgically fitted at the distal ileum following the procedure described by Sauer et al. (1983).Post-surgery, pigs were housed individually in plexi-glass pens adjacent to one another with supplemental heat for recovery.A six-day recovery period followed post surgery prior to the start of the experimental periods.The surgical incision site was cleaned twice daily with warm water and zinc oxide cream was applied until the incision was healed.Commercial feed and water were provided ad libitum during the recovery period.Throughout the duration of the study, the cannula site was cleaned twice daily with warm water and zinc oxide was applied to prevent skin irritation.(Xin et al. 2022), lentil (Wang and Daun 2006), and pea (Nosworthy et al. 2021).
pelleting a high corn starch-based diet.The pulse samples were ground via a hammer mill (G.J. Vis Model: VISHM2014) through a 3.97 mm screen.Pulse experimental diets were steam-pelleted (UAS-Muyang Model: MUZL350II) at 80 • C. The pelleter had a retention time of 45 s with the experimental diets passing through a 4 mm die.Hot pellet temperatures exiting the die were 78, 78, and 80 • C for faba bean, lentil, and yellow pea, respectively, prior to cooling.Analysed chemical compositions of experimental diets are reported in Table 3.
Pigs were randomly assigned to experimental treatments based on a 4 × 4 replicated Latin square design, where each pig received each treatment in only one period.Experimental periods were 11 days in duration, with dietary adaptation occurring from days 0-6 and ileal digesta collection occurring on days 9 and 10.Daily feed rations were offered as two meals at 0830 and 1630 h, with water added to the feed to create a wet mash.Feed intake (kcal/day) was calculated based on 2.8× maintenance digestible energy (DE) requirement (NRC 2012).Any feed refusals were collected and weighed post meal.

Sample collection, processing, and analyses
Ileal digesta was collected on days 9 and 10 of each period between 0830 and 1630.Briefly, plastic bags containing 10 mL of 10% formic acid (to inhibit microbial activity/growth) were attached to the opening of the cannula with elastic bands.Bags were checked periodically and replaced when full.Collected digesta was placed in a fridge at 4 • C until the end of collection.At the end of the collection period, digesta was pooled per pig per period, divided into aliquots, and stored at −20 • C until analysis.
Prior to chemical analyses, ileal digesta samples were freeze dried and finely ground through a 1 mm screen.Exper-imental diets and ileal digesta were analyzed for proximate analysis, AA, and TiO 2 .Dry matter was determined following AOAC method 930.15 (AOAC 2005), with slight modifications.Six to eight grams of feed samples were dried in the oven at 135 • C for 3 h while 2 g of fecal/digesta samples were dried in the oven at 135 • C for 3 h.Ash was determined based on AOAC method 942.05 (AOAC 2005) with modifications as described by Tansil et al. (2022).Nitrogen content of samples was determined using a LECO analyzer (LECO Corporation, St. Joseph, MO; AOAC method 968.06;AOAC 2005).Crude protein was calculated by multiplying the nitrogen content by a factor of 6.25.Crude fat was determined via the petroleum ether extraction method (AOCS Am 5-04; ANKOM Technology, 2009).Neutral and acid detergent fiber, following the filter bag technique ANKOM A200 and A2001.Titanium content was analyzed following Myers et al. (2004) with minor adaptations as described by Tansil et al. (2022).AA content was determined in feed and digesta via sample hydrolysis and ultra-performance liquid chromatography (Waters Corporation, Milford, MA, USA) analysis.Samples were prepared and analyzed for indispensable and dispensable AA following Cargo-Froom et al. (2022).

Statistical analysis
Data were analyzed using a mixed model in SAS (Version 9.4; SAS Inst.Inc., Cary, NC) via Proc Glimmix, where diet was treated as a fixed effect and pig and period as random effects.
Least squared means were compared using Tukey HSD.Significance was determined when P ≤ 0.05.

Results
All pigs on trial recovered well from surgery and remained healthy throughout the entire study.One pig was removed from study and euthanized on day 10 of the final period due to loss of cannula at the start (first hour) of the sampling period.Data for period 4 resulted in an n = 7 for the yellow pea experimental diet instead of 8.
AIDs of proximate nutrient content, acid detergent fiber, neutral detergent fiber, and AA and endogenous losses are reported in supplementary Tables 1 and 2, respectively.SID values for CP and AA are reported in Table 4.There were no differences in the SID of CP when the three steam-pelleted experimental treatments were compared (P > 0.05).With the exception of Trp, glycine (Gly), and proline (Pro), all AA had digestibility values of 80% or greater.There were no differences in SID of indispensable and dispensable AA when the three steam-pelleted experimental treatments were compared (P > 0.05).
The DIAAS values for the test ingredients are reported in Table 5.The DIAAS values were 30, 58, and 27 for steampelleted faba bean, lentil, and pea, respectively.Tryptophan

Discussion
There were no differences in SID of CP and AA across the three test pulses.Although the CP content of the experimental diets were lower (∼10% CP) compared to the target CP content of other ileal digestibility studies (∼18% CP; Mariscal-Landín et al. 2002;Friesen et al. 2006;Stein and Bohlke 2007;Jezierny et al. 2010;Presto et al. 2011;Eklund et al. 2012;Nørgaard et al. 2012;Hugman et al. 2021aHugman et al. , 2021b)), due to similarity in CP content (∼10%) across the experimental pulse diets, the SIDs of CP and AA of the investigated pulses are still comparable.Additionally, Zhang et al. (2005) determined that the SID of AA in soybean protein diets should be measured at a CP content of 10%-20%.Although soybean is not categorized as a pulse, both soybean and pulses are legumes, act as a source of vegetable-based protein, and have similar properties (e.g., nitrogen fixation, cardiovascular risk prevention, etc.; Vollman 2016).
Results demonstrated that none of the pulses evaluated were superior for inclusion in swine diets as a source of protein.The SID values reported showed that, with the exception of Trp, faba beans, lentils, and yellow peas provide a readily digestible source of indispensable AA when included in the diets of growing swine.While the SID for Trp was lower than the other AA reported, it also had the highest variability for all AA between analyzed samples.The lower SID of Gly and negative SID of Pro are most likely attributed to endogenous losses that were not captured by the NFD diet, as these AAs tend to be abundant in endogenous contributions (Thewis et al. 1996;Stein et al. 1999).
The SID values of CP for faba beans, lentils, and yellow peas fall within the range of values currently reported in literature for other plant-based protein sources (Mariscal-Landín et al. 2002;Friesen et al. 2006;Stein and Bohlke 2007;Jezierny et al. 2010;Presto et al. 2011;Eklund et al. 2012;Nørgaard et al. 2012;Han et al. 2020;Hugman et al. 2021aHugman et al. , 2021b)).The SID values of faba beans for indispensable AA were higher, with the exception of Arginine (Arg) and Trp, than data reported in literature for raw faba beans (Mariscal-Landín et al. 2002;Jezierny et al. 2010;Presto et al. 2011), with up to a ∼26% unit difference in digestibility values.Tryptophan digestibility was similar to data reported by Jezierny et al. (2010; average digestibility 61%) in raw faba beans, whereas compared to data reported by Mariscal-Landín et al. (2002), digestibility was ∼7%-21% units lower.
There are little data reported for the SID of lentils; however, for the majority of indispensable AA investigated, our study reported higher SID values than those reported in literature (Woyengo et al. 2014;Hugman et al. 2021a).Tryptophan is reported to be 10%-34% units lower in lentils in this study compared with SID of raw and processed (steam pelleting and extrusion) lentils in literature (Woyengo et al. 2014;Hugman et al. 2021a).
Field peas have been the most extensively studied pulse in swine literature, and, thus, there is substantially more data for comparison with other pulses such as beans, chickpeas, lentils, etc.The SID of Arg from yellow peas was similar to those reported in raw pulses but was lower than those reported for extruded pulses (Mariscal-Landín et al. 2002;Friesen et al. 2006;Stein and Bohlke 2007;Jezierny et al. 2010;Eklund et al. 2012;Nørgaard et al. 2012;Han et al. 2020;Hugman et al. 2021b).Extrusion of field peas resulted in higher SID of AA compared to values reported by this study, with the exception of Met, which had SID values higher or comparable to extruded peas (Mariscal-Landín et al. 2002;Stein and Bohlke 2007;Hugman et al. 2021b).Compared with raw/whole samples, digestibility of AA from peas in this study was comparable or higher than data reported in literature for field peas (Mariscal-Landín et al. 2002;Friesen et al. 2006;Jezierny et al. 2010;Eklund et al. 2012;Nørgaard et al. 2012;Han et al. 2020).
Traditionally, Met and Trp are limiting indispensable AA in pulses.In the present study, regardless of pulse type, Met digestibility was 3%-36% unit greater than previously reported values of both raw and heat processed pulses (Mariscal-Landín et al. 2002;Friesen et al. 2006;Jezierny et al. 2010;Presto et al. 2011;Eklund et al. 2012;Nørgaard et al. 2012;Han et al. 2020;Hugman et al. 2021aHugman et al. , 2021b)); however, the total amount of Met supplied by the pulses was not enough to meet the Met requirement of growing pigs (NRC 2012).Although the SID of Trp reported in this study were low, Trp digestibility also had the highest variability of all AA.The low Trp digestibility of all pulse diets could be in part due to the variability in Trp analysis.Similarly to Met, the total amount of Trp supplied by the pulses was not enough to meet the Trp requirement of growing pigs (NRC 2012).These results suggest that these varieties of faba beans, lentils, and field peas alone cannot meet the indispensable AA requirements of growing swine but could be a source of dietary protein when combined with an additional protein source or crystalline AA supplementation in the diet, potentially sparing some common protein supplying ingredients that are in high demand/cost (e.g., soybean).
Aside from Trp, some possible explanations for greater SID values of the test pulses compared to previous values in this study could be due to effects of processing.In this study, the experimental pulse diets were steam-pelleted at 80 • C. Processing can improve digestibility either through the reduction of ANF or via denaturation of proteins.While Hugman et al. (2021a) reported an increase in SID for CP and all AA when lentils were steam pelleted at 80-85 • C (which could be due to denaturation of proteins), Stein and Bohlke (2007) only noted significant increases in SID for CP and AA of extruded pulses (75-115 • C) but not pelleted (75 • C).As such, further studies on the effects of processing on SID of CP/AA, in addition to the digestibility of raw pulses, should be further investigated, as feed processing and feeding practices can vary depending on farm/location (e.g., mixed meal versus pelleted diets).This will provide more accurate values for the use of raw or processed pulses in animal feed.
The DIAAS is considered to be the most accurate metric to determining the protein quality of food for humans.Given that determination of DIAAS requires ileal digestibility values for all IAA, which are difficult to acquire from humans, ileal digestibility values from animal models are often used.Swine are considered the gold standard model for CP and AA digestibility for human nutrition (Bergen 2007;Guilloteau et al. 2010;Deglaire and Moughan 2012), where the FAO has indicated that growing pigs are an ideal model establishing ileal digestibility values for DIAAS calculations for human protein foods.As such, both swine and human nutrition can gain valuable insight and data to be used by both industries via determining the SID of AA from pulse ingredients in swine.
Upon the introduction of the DIAAS method, it was suggested that, if implemented into a regulatory framework, "source of" protein claims could only be made for foods and ingredients that have a DIAAS score of 75-99 (good source) or ≥100 (excellent source); any value below a score of 75 cannot make a protein source claim (FAO 2011b).While DIAAS has not been implemented in any regulatory framework to date, results from this study demonstrated that, under the DIAAS framework, faba beans, lentils, and peas would not be eligible for a protein content claim for human food.This was due to the low AA score from Trp.However, as previously discussed, Trp SID values in this study showed considerable variability between animals.Additionally, comparison of Trp digestibility values within the published literature can be challenging.Some studies do not report Trp content or digestibility values of pulses and (or) only report the digestibility of processed but not raw pulses, which can make comparison of data and identifying trends more challenging (Friesen et al. 2006;Ma et al. 2011;Eklund et al. 2012;Nørgaard et al. 2012;Nosworthy et al. 2017Nosworthy et al. , 2018aNosworthy et al. , 2018b)).Other studies also determined that pulses would not be eligible for a protein claim under the DIAAS framework.Similar results were demonstrated by Han et al. (2020), using the pig as a model, in a variety of cooked pulses including faba beans (identified as broad beans in the Han et al. study) and peas and determined the DIAAS score of broad beans to be 60 (leucine) and that of peas to be 68 (lysine).Using rats, DIAAS for cooked (extruded, boiled, and baked) faba beans, lentils, and peas that were fed cooked pulses, ranged between DIAAS of 54-61 (SAA) for faba beans, 44-58 (SAA) for lentils, and 64-70 (SAA) for peas (Nosworthy et al. 2017(Nosworthy et al. , 2018a(Nosworthy et al. , 2018b)).However, rather than ileal digestibility, true fecal digestibility values were used in the DIAAS calculations.
Currently, it appears that these pulses do not qualify for protein claims based on DIAAS calculations.However, arbitrary cut-off score of ≥75 for identifying a protein food under the DIAAS framework has also garnered some scrutiny (Marinangeli and House 2017).Moreover, from a human dietary perspective, given the variety of foods that comprise diets, the protein quality of dietary patterns, which would include the concept of complementarity, would be a valid consideration (Craddock et al. 2021).Nevertheless, further investigation is warranted into other pulses/varieties for protein claims in relation to pulses for human nutrition.
There were limitations to this study, which further research should attempt to answer.This study did not measure differences between raw and processed pulses, ANF content, or the differences in ANF content between raw and processed pulses.Additionally, this study only evaluated one processing parameter, which was based off common industry practice.Some pulses, such as peas, have many varieties within their category, and variety was not evaluated.Based on reported literature, variety (e.g., amino versus Australian peas) can influence the CP and AA digestibility of pulses (Mariscal-Landín et al. 2002;Friesen et al. 2006;Jezierny et al. 2010).Consequently, comparison across varieties within a pulse category should be considered.

Conclusion
Results from this study suggest that pulses may be a suitable protein supplying ingredient to include in the diet of growing swine.The investigated pulses provided a readily digestible source of protein and AA when steam-pelleted.Methionine and Trp were not supplied by the ingredients in sufficient amounts to meet requirements.Thus, if pulses are included in swine feed, other AA supplying sources (crystalline AA or other protein sources) should be included to ensure that AA requirements are being met.The DIAAS values of these ingredients were determined with no protein claim able to be made.Ultimately, by quantifying these different values (SID and DIAAS), the protein quality of these pulses can be better understood for use in swine feed and consumption as humans.

Table 1 .
Chemical analyses of test pulses, dry matter basis.

Table 2 .
Experimental diet formulations, as fed basis.

Table 3 .
Analyzed chemical composition of experimental diets, dry matter basis.
where x is the individual indispensable AA (except Met and Phe) and total SAA (Met + Cys)/total aromatic AA (Phe + Tyr) of interest, IAA x is the indispensable AA concentration in the treatment diet of interest (faba bean, lentil, and pea; mg/g of CP), and SID x is the SID coefficient of individual indispensable AA (except Met and Phe) and the average SID coefficient of Met + Cys and Phe + Tyr.

Table 4 .
Standardized ileal digestibility (%) of crude protein and amino acids of faba bean, lentil, and yellow peas fed to growing pigs.

Table 5 .
Digestible indispensable amino acid score of faba bean, lentils, and yellow peas.