Thermal processing methods differentially affect the protein quality of Chickpea (Cicer arietinum)

Abstract Chickpea is a widely produced pulse crop, but requires processing prior to human consumption. Protein bioavailability and amino acid quantity of chickpea flour can be altered by multiple factors including processing method. For this reason, the protein quality of processed chickpea flour was determined using in vivo and in vitro analyses for processed chickpeas. Processing differentially affected the protein digestibility‐corrected amino acid score (PDCAAS) of chickpeas with extruded chickpea (83.8) having a higher PDCAAS score than both cooked (75.2) and baked (80.03). Interestingly, the digestible indispensable amino acid score (DIAAS) value of baked chickpea (0.84) was higher compared to both extruded (0.82) and cooked (0.78). The protein efficiency ratio, another measure of protein quality, was significantly higher for extruded chickpea than baked chickpea (p < .01). In vivo and in vitro analysis of protein quality were well correlated (R 2 = .9339). These results demonstrated that under certain circumstances in vitro methods could replace the use of animals to determine protein quality.

acid content, in conjunction with reduced protein digestibility compared to other protein sources, has been implicated as the main reason for the lower nutritional value of chickpea protein (Mudryj, Yu, & Aukema, 2014). Anti-nutritive factors such as trypsin inhibitors, protease inhibitors, and lectins can alter amino acid bioavailability by limiting protein digestibility (Tavano et al., 2008). Preparatory methods such as extrusion, baking, or cooking can alter the concentration and/or activity of these anti-nutritive factors and may thereby alter the bioavailability and digestibility of amino acids and protein, respectively, in chickpea flours.
Protein quality can be determined by multiple methods including protein efficiency ratio (PER), protein digestibility-corrected amino acid score (PDCAAS), and digestible indispensable amino acid score (DIAAS). PER is a measurement of growth mandated for use in the regulation of Canadian protein content claims (Health Canada, 1981). In the United States, PDCAAS is the required method of determining protein quality for claim purposes (FAO/WHO, 1991) while the most recently developed method for measuring protein quality, DIAAS, is not used for regulatory purposes in any jurisdiction (FAO/WHO, 2013). One aspect of the current study was to determine the effects of extrusion, baking, and cooking (boiling) on the protein quality of chickpeas. Protein digestibility was also determined via in vitro methodology for the calculation of in vitro PDCAAS, which was used to investigate the correlation between in vivo and in vitro methods of protein quality assessment (Nosworthy, Franczyk, et al., 2017). This study also afforded an opportunity to investigate the potential for cultivar or growing location to impact protein content and amino acid composition of Canadian chickpeas, as had been previously demonstrated in India (Singh, Kumar, & Gowda, 1983).

| Statement on animal ethics
All animal procedures received approval by the University of Manitoba's Institutional Animal Care Committee, which utilize the appropriate guidelines established by the Canadian Council on Animal Care (CCAC, 2017).

| Sample procurement and preparation
Samples of chickpeas for processing were provided by Saskcan Pulse Trading, Thompsons Ltd., and Viterra. Chickpeas from different suppliers were combined and thoroughly mixed before processing.
Milling of the combined samples to generate flour for extrusion and baking was performed milled on a hammer mill using a 0.050 inch screen (Jacobson 120-B hammer mill) (Nosworthy, Franczyk, et al., 2017). The extrusion and baking of the chickpea flour, as well as the cooking of the chickpeas, were performed as previously described . Baked samples underwent hammer milling (Fitz mill-model #D comminutor VHP-506-55B), with screen hole size of 0.020 inch, round, followed by a 20 mesh screening on a sifter

| Sample analysis
Percent crude protein (CP; N × 6.25) was determined via Dumas Nitrogen Analyzer (Dumatherm DT, Gerhardt Analytical Systems), while percent dry matter (DM) and ash were determined according to AOAC guidelines (AOAC, 1995). The selection of a Jones factor of 6.25 was done according to recommendations for the determination of protein quality (AOAC, 1995). A control sample (NIST 3234, National Institute of Standards and Technology) was included in each amino acid assay to ensure the accuracy of the assay. Percent crude fat was determined by hexane extraction and gravimetrics (AOAC 2003.06). Sulfur amino acid content was determined according to AOAC 994.12 with the remaining amino acids, excepting tryptophan, determined according to AOAC 982.30. Analysis of tryptophan was performed as previously described (ISO, 2005;Nosworthy, Franczyk, et al., 2017).

| Statistics
True fecal protein digestibility (TFPD) and PER results (n = 10) were compared via one-way ANOVA with Tukey's selected as the post hoc test. Correlations between both in vivo and in vitro digestibilities, and PDCAAS/in vitro PDCCAS (n = 4) were determined via regression analysis (GraphPad Prism, 7.0, GraphPad Software).

| Proximate analysis
Sample proximate data are presented in Table 1, with crude fat/ protein and amino acid composition being presented on a DM basis.
While the dry matter of the unprocessed chickpea flour (91.95%) was lower than that of any processed flours, it was similar to previously reported results (92.32%) (Canadian Nutrient File, 2015).

| Amino acid score and protein digestibility
Sample amino acid composition is presented in Table 1, and amino acid scores are presented in Table 2. The first limiting amino acid in all processed chickpea flours was tryptophan. While this agrees with certain studies (Nosworthy, Franczyk, et al., 2017;Tavano et al., 2008), others have found that chickpeas were initially limited in the sulfur containing amino acids, cysteine, and methionine (Jukanti et al., 2012;Wang & Daun, 2004). The amino acid scores of the extruded (0.97) cooked (0.86) and baked (0.95) chickpeas were higher than anticipated. Compared to previous work in chickpeas, which found amino acid scores of 0.61-0.62, the chickpeas used in this study have a different composition that is more similar to the human nutritional pattern for children aged 2-5 years put forth by the FAO/ WHO (1991), resulting in a higher amino acid score (FAO/WHO, 1991). It is also worth noting that while chickpeas can be limiting in sulfur amino acids, the amino acid score for methionine + cysteine is either the same, 1.03 for extruded flours, or greater, 1.07 for baked flours, than that found in casein (1.03). These high amino acid scores for chickpea sulfur amino acids have been corroborated by similar findings in other chickpea samples (Bai, Nosworthy, House, & Nickerson, 2018), and the fidelity of the amino acid protocol has been confirmed via the use of standards and the analysis of a control sample, soy flour. The difference in amino acid composition and the resulting amino acid scores of these processed chickpeas could be due to the differences in varieties, crop growing location, or other environmental factors. Determining how agronomy can influence amino acid quantity could potentially result in higher quality proteins from plant-based sources.
Protein digestibility values as determined by in vitro and in vivo measurement are presented in Table 3. Chickpea TFPD significantly differed between cooked (87.17%) and baked (84.62%; p < .05).
No difference was detected among cooked, baked, and extruded (86.56%) samples. While these digestibilities are higher than reported for heated chickpea flour, 78.75% (Tavano et al., 2008), the cooked true protein digestibility is similar to that found in another study, 85.02% . The chickpea protein digestibilities found in this study are similar to previous findings in canned chickpeas, 88%-89% (FAO/WHO, 1991). The digestibility of raw chickpeas, as determined in vitro, has been reported as between 34%-76% (Jukanti et al., 2012), with one study determining a protein digestibility of 89.01%, which increased to 96.94% after heating (Monsoor &Yusuf, 2002). This variability in in vitro digestibilities could be attributed to different methods of analysis as methods can differ in number/type of digestive enzymes, pH, and incubation time, all of which can alter the final value attributed to protein digestibility. In this study, in vivo and in vitro protein digestibility values differed in that while baked chickpea had the lowest digestibility in vivo and in vitro, cooked chickpea had the highest in vivo digestibility, while extruded chickpea had the highest digestibility in vitro.
For cooked and baked chickpeas, in vitro digestibility was lower than in vivo, while the in vitro digestibility was greater than in vivo for extruded chickpeas. As the in vitro method used in this study incorporates a limited representation of the digestive process compared to an in vivo system, it is unsurprising that this in vitro system would not perfectly mimic the digestive process.

| PDCAAS and in vitro PDCAAS
The protein digestibility-corrected amino acid score (PDCAAS) and in vitro PDCAAS score are presented in study used a one-step pH drop method (Tinus et al., 2012), and for determining in vitro protein digestion for comparison with that found in the rodent model. The correlation between in vitro and in vivo digestibility was R 2 = .7344, while the correlation between PDCAAS and in vitro PDCAAS had an R 2 value of .9339 (p = .0336). This relationship between in vivo and in vitro protein quality is similar to that found in other plant-based protein sources (Nosworthy, Franczyk, et al., 2017;Tavano et al., 2016), further supporting the concept of using in vitro PDCAAS as a surrogate for animal experimentation (Figure 1).

| DIAAS
The digestible indispensable amino acid score (DIAAS) was recom-  Table 4. The DIAAS value for baked chickpea (0.84) was higher than cooked (0.78) and extruded (0.82), and for all processing methods, sulfur amino acids were limiting. Previous work found a DIAAS value of 0.67 for cooked chickpeas , whereas in this study the DIAAS value for cooked chickpea was 0.78. When compared to PDCAAS, DIAAS values for baked and cooked were higher, while the DIAAS value for extruded chickpea was lower. This might be explained as the reference pattern used in the determination of DIAAS and PDCAAS is different; specifically, the requirements for sulfur amino acids were lowered to 25 mg/g protein for DIAAS from 27 mg/g protein for PDCAAS, and the tryptophan requirement from was reduced from 11 mg/g (PDCAAS) to 8.5 mg/g (DIAAS).

| PER
Compared to PDCAAS and DIAAS, the protein efficiency ratio (PER) is a growth measurement comparing weight gain over a period of 28 days to the amount of protein consumed. Currently, Health Canada mandates the use of PER as a protein quality measurement to regulate content claims for protein (Health Canada, 1981). The chickpea PER data are presented in Figure 2. Extruded chickpea had a significantly higher PER than baked (p < .01); however, no significant difference was found between either extruded and cooked or cooked and baked chickpeas. A study investigating baked chickpeas determined a PER of 2.88, compared to 2.3 in this study while previous work on cooked chickpeas determined a PER of 2.32 versus 2.42 in this study (Nosworthy, Franczyk, et al., 2017;Tavano et al., 2008).
To account for measurement variability, PER values are also adjusted to the relative PER for the control, casein, which is set to 2.5 (Health Canada, 1981). These values, presented in Table 3, indicate that for chickpeas, extrusion resulted in the highest growth rate based on protein consumption, followed by cooking and baking.

| Effects of cultivar and location on proximate and amino acid composition of chickpeas
The proximate and amino acid composition of three chickpea cultivars (Frontier, Leader, and Orion) grown at three locations in Saskatchewan (Cabri, Limerick, Moose Jaw) in 2014 are presented in Table 5 with the resulting amino acid scores presented in Table 6.

| CON CLUS ION
In summary, processing is capable of altering protein quality through changes in either the amino acid composition or protein digestibility. This study has demonstrated that extrusion is the optimal method for producing the product with highest protein quality, while for home preparation baking chickpeas would provide a higher protein quality than cooking. The method of determining in vitro protein digestibility used in this study resulted in a good correlation between in vivo and in vitro measurements of PDCAAS, providing more support to the use of in vitro methods for determining protein quality. An overview of protein content and amino acid composition for three chickpea cultivars also revealed potential variation between protein content and amino acid scores depending on growing locations, suggesting that further study of the effects of environment x genetic interaction on protein content and quality can be pursued.

ACK N OWLED G M ENTS
Funding for the study was provided via Agriculture and Agri-Food