Abstract:
163 In vitro assessment of starch digestion in pigs R.J. van Barneveld1 and M.R. Zarrinkalam2 1 2 Barneveld Nutrition Pty Ltd and BECAN Consulting Group, 19_27 Coonan Rd, South Maclean Qld 4280 Office of Industry Liaison, Adelaide University Research Park, Thebarton Campus, 35_37 Stirling St, Thebarton SA 5031 rob@barneveld.com.au Summary Application of in vitro feed evaluation assays that simulate digestion in the pig have limited potential for routine application at the feedstuff producers site of ingredient receival. If the assays are to have a role, it will be the strategic assessment of grain characteristics that may influence digestion in the small intestine of pigs and subsequent nutrient yield. Starch composition and factors affecting access to starch within different grains may have a significant influence of energy yield, but it is difficult to detect these differences using traditional twostep enzymic procedures based on pepsin and pancreatin. An in vitro assay that focuses on starch digestion alone is likely to be more sensitive to how differences in starch composition and starch access may influence energy digestion. This paper discusses the application of the glucose release index as a means of assessing factors that may influence starch digestion in vitro, and how factors such as grain hardness and the protein matrix surrounding starch granules in cereals may be influencing starch digestion in the small intestine in pigs. Results suggest that the protein matrix in barley presents less of a physical barrier to starch digestion than in wheat and sorghum and that roller milling may be a far more effective way of preparing sorghum for use in nonpelleted pig diets. Keywords: pigs, starch digestion, in vitro assays, glucose release index pig feeds since 1995 (Boisen 2000). In that case the methods have been implemented to finetune energ y and amino acid supply according to predicted actual requirements defined using pig growth models. The shortcomings of this approach include: the in vitro models utilised are difficult to apply at the point of grain receival due to the time taken to complete the analysis; given the difficulty in simulating the dynamic processes that occur in the gastrointestinal tract of an animal, it is somewhat unrealistic to expect a simplistic laboratory assay to adequately define the nutritional quality of all ingredients (Moughan 1999); in the case of pigs, simulation of whole tract energy digestion assumes that digestible energy is the most reliable measurement of available energ y for the animal. Introduction Application of in vitro feed evaluation assays that simulate digestion in the pig have limited potential for routine application at the feedstuff producers site of ingredient receival. If the assays are to have a role, it will be the strategic assessment of grain characteristics that may influence digestion and subsequent nutrient yield. The general application of in vitro methods for the analysis of feed ingredient quality has been widely promoted by countries such as Denmark where such methods have been in general use for the evaluation of Based on criteria of speed, ease of use, accuracy, and cost it is difficult to envisage any form of routine feed ingredient analysis exceeding the current potential near infrared spectroscopy (NIRS). Application of NIRS is already widely accepted as the basis for grain classification and payments using measurement of parameters such as crude protein and moisture and potential already exist to measure in vivo parameters such as pig digestible energy (DE; van Barneveld et al. 1999). In addition, recent research and development undertaken in the Premium Grains for Livestock Program involving grains, livestock and feed milling sectors suggests that NIRS calibrations are capable of simultaneously measuring a suite of chemical parameters. These include proximates, antinutritional factors, amino acids, selected soluble and insoluble non starch polysaccharide components, starch characteristics and fatty acids plus a range of in vivo measurements such as ileal and faecal digestible energy for pigs, or apparent metabolizable energy for broiler chickens and laying hens, and may soon be available for routine use Recent Advances in Animal Nutrition in Australia, Volume 14 (2003) 164 van Barneveld, R.J. and Zarrinkalam, M.R. (R.J. van Barneveld, unpublished). NIRS also holds significant potential for realtime analysis of grain composition that may flag a need to adjust diet formulations and consider strategic application of feed additives such as exogenous enzymes. That may not be so for routine application of in vitro assays based on wet chemistry unless they can be applied for determinations not suited to NIRS. In broad terms, a nutritional quality definition system is adequate if it provides a means of discrimination between ingredients, and reflects the level of accuracy that can influence commercial responses in a livestock production system (van Barneveld 2001). There will be improvements in the ability to separate the potential supply of net energ y from similar grain types if techniques increase the focus on energy digestion in the small intestine of the pig. An understanding of those factors that have the greatest influence on digestion in the small intestine will increase our capacity to manipulate these grain characteristics and it is here that in vitro assays may have a very useful strategic role. The objective of this paper is to demonstrate the potential for in vitro assays to identify factors that may influence energy supply from grains in the small intestine of pigs and thus provide a basis for more detailed assessment using in vivo measurements. The glucose release index will be used as a specific example of strategic application of an in vitro assay for this purpose. fatty acid type and content; starch composition; protein matrix; grain hardness; and lignin bound to phenols and proteins. If we focus specifically on energy supply from the small intestine of pigs, the starch composition and the access to starch granules within different grains may have a significant influence of energy yield, but it is difficult to detect these differences using traditional two step enzymic procedures based on pepsin and pancreatin (Figure 1). An in vitro assay that focuses on starch digestion alone is likely to be more sensitive to how differences in starch composition may influence energy digestion in the small intestine of pigs. The `glucose release index' as an in vitro assay for starch digestion Several in vitro starch digestibility assays have been developed and used across a variety of species including humans, ruminants and poultry (Zarrinkalam 2002). The basic principle applied to most in vitro starch digestibility assays is similar and aims to partially mimic the digestive conditions present in the small intestine by using similar starch digestion enzymes and pH conditions. Englyst et al. (1992) used a more rapid approach to measure the digestibility of starch and other carbohydrates such as fructose by designing a single time point multienzyme assay. This assay has been successfully used for rapidly assessing variations in starch digestibility between different foodstuffs (Englyst et al.1996). A starch digestion index was utilised and defined as the percentage of rapidly digested starch (following a 20 minute incubation) to total starch (Englyst et al. 1992). Zarrinkalam (2002) refined this methodology specifically for cereal grains. Only two digestive enzymes (heat stable amylase and amyloglucosidase; Figure 1) out of the 6 enzymes originally used by Englyst et al. (1992) were considered necessary to investigate the variation in starch digestibility between cereal grains, since almost all of the digestible carbohydrate in cereal grains is starch. Data from this modified procedure of determining the ratio of the released glucose in the initial stage of starch digestion to the released glucose at the completion of starch digestion was defined as the glucose release index (GRI). The procedure accounts for the fact that differences in the amount of glucose released from the digestion of starch in grains are maximal during the initial rapid phase of starch digestion. Restricting measurements of the progression of an enzymic reaction to a period where less than 2050% of total substrate consumption has occurred, provides a value that reflects the conversion rate of a substrate (starch) to its product (glucose). The basis for strategic application of in vitro assays Moughan (1999) and Boisen (2000) described traditional approaches to simulation of precaecal and total tract digestion in pigs. Most of these in vitro systems are based on twostep (in the case of precaecal digestion) or threestep (in the case of total tract digestion) enzymic procedures (Figure 1). Comparative application of these procedures by Boisen (2000) demonstrates clearly the lack of sensitivity of whole tract digestion versus precaecal digestion (Figure 2) when assessing ingredients varying widely in composition. However, the capacity of twostep and threestep enzymic procedures applied by Boisen (2000) to discriminate between similar grain types is not clear. While the Boisen (2000) assays focus on simulating digestion, it is difficult to apply them in a way that allows an assessment of how specific chemical and physical characteristics of the grain may influence digestion. Black (2001) suggested that differences in the nutritional quality of grains could be largely attributed to combinations of the following factors: gross chemical composition; cell wall quantity, structure and composition; In vitro assessment of starch digestion in pigs 165 Pre-caecal (two_step) Sample Total tract (three-step) Pepsin, pH 2.0 4 hours Pepsin, pH 2.0 6 hours Pancreatin, pH 6.8 4 hours Pancreatin, pH 6.8 18 hours, 39�C Viscozyme, pH 5.0 18 hours, 39�C Filtration Undigested residue Non-fermented residue Ileal DM Ileal N Faecal DM Faecal N `Glucose Release Index' - amylase, pH 7.0 3 minutes, 50�C amyloglucosidase 18 minutes, 40�C Glucose release Figure 1 Flow diagram depicting traditional two_step and three_step in vitro incubations and the comparative application of the `glucose_release' assay. 166 van Barneveld, R.J. and Zarrinkalam, M.R. Ileal 100 90 80 70 60 50 40 30 20 10 0 Faecal In vitro digestibility (%) Figure 2 In vitro digestibility of dry matter of common feedstuffs after simulation of pre_caecal and whole tract digestion (Boisen 2000). Grain factors influencing starch digestion that can be assessed using the GRI Wheat and sorghum will be used as example cereals to discuss grain factors influencing starch digestion in pigs, due to their significantly different cell structure and configuration. Pigs appear to digest similar amounts of energy from cereal grains such as wheat and sorghum in the small intestine, but the lower starch and fat content in wheat may mean that enzymic access is superior with a higher energy yield per unit of those substrates (van Barneveld et al. 2001). Examination of the cellular characteristics of wheat and sorghum may identify some of the reasons for differences in the efficiency of starch digestion in cereals, within and between species. Wheat consists of elongated cells containing a mixture of twothirds large, lenticular starch granules (830 �m; A granules) and onethird near spherical granules of <8 �m in diameter (B granules). In contrast, sorghum is characterised by small, tightly packed, angular cells containing uniformsized starch granules encapsulated in a protein matrix. It is reasonable to suggest that the capacity to break these sorghum cells and disrupt the associated protein matrix may have a significant influence on starch digestion. The elongated cells of wheat are likely to be easier to rupture through chewing or processing prior to feeding, making the cell contents more accessible to enzyme degradation, whereas the nature of the cells in sorghum means they are more likely to remain intact following chewing or processing. To this end, assessment of the potential influence of the protein matrix surrounding starch cells and grain hardness, respectively, may provide some insight into how the efficiency of starch digestion may be improved in the small intestine of pigs. Ba rle Suga y r be et pul M p ai ze fee d m eal Wh eat br an Ra pe se ed m ea l Gra ss m eal meal So ybe an Wh eat Ingredients Application of the GRI to assess the influence of protein matrix on starch digestion The influence of protein matrix on the GRI of starch was determined by digesting grain protein with pepsin at pH 2 for 60 minutes at 39�C prior to conducting the GRI assay (Zarrinkalam 2002). The use of pepsin protease, and the pH and incubation conditions were chosen to simulate the digestive conditions present in the stomach of monogastric animals. Crude protein content of the grain was not related to t he i nfluence of t he p rotein m atrix on GRI (Figure 3). There were different responses between grain types to pretreatment with pepsin which has an ability to remove protein matrix (as revealed by scanning microscopy). Pretreatment of barley with pepsin resulted in a consistent decrease in GRI compared with wheat and sorghum where increases in GRI were observed in all cases. Zarrinkalam (2002) suggested that the protein matrix in barley presents less of a physical barrier to starch digestion than in wheat and sorghum and that the protease activity may actually impede starch digestion in barley. This conclusion is further supported by the result that non pepsintreated barley grains exhibited the highest average GRI value compared to sorghum and wheat. Also, in sorghum, despite a significant increase in the average GRI value after protease treatment, this value was still significantly lower than in barley and wheat samples treated with pepsin. This may be related to the lower surface area to volume ratio of starch granules in sorghum, which could restrict digestive enzyme accessibility to starch. Another important difference between sorghum and other grain types such as barley and wheat is its prolaminerich In vitro assessment of starch digestion in pigs 167 4 Difference in GRI (pepsin-nonpepsin treatment) 3 2 Differences Crude protein (%, DM) 20 18 16 14 1 0 12 10 -1 -2 -3 -4 8 6 4 2 0 Cultivars Figure 3 Difference between the glucose release index of untreated barley, wheat and sorghum and pepsin pre_treated barley, wheat and sorghum (Zarrinkalam 2002). protein matrix ( and kaffarins). The protein matrix in sorghum, particularly in the peripheral area of the corneous endosperm region, may not have been completely digested by the pepsin treatment, restricting enzyme accessibility to starch granules. The above results are supported by in vivo responses by pigs to application of exogenous enzymes containing an active protease (Figure 4). The ileal:faecal DE ratio of wheat fed concurrently with an exogenous enzyme was significantly improved relative to barley and sorghum, possibly due to an increase in energy yield from starch in the small intestine (R.J. van Barneveld, unpublished data). In this instance the sorghum sample tested was a waxy sorghum (which was not tested using the GRI), and hence the response to an exogenous enzyme may have been atypical. Application of the GRI to assess the influence of grain hardness on starch digestion Zarrinkalam (2002) applied the GRI to investigate the influence of milling process on starch digestion in barley. The GRI values determined for samples milled through a 2 mm screen and samples rollermilled were subtracted from the corresponding GRI values of samples milled through a 0.5 mm screen and the dif ference expressed as a proportion of the corresponding 0.5 mmmilled value (Figures 5 and 6). Zarrinkalam (2002) reported a significant (P<0.05) linear relationship between the grain hardness index (determined on whole grain using single kernel analysis; Bread Research Institute, North Ryde), and the change in GRI that occurred when grains were either milled through a 2 mm screen or roller milled relative to the 0.5 mmmilled samples. As the grain hardness increased, the GRI generally decreased when the grains were roller milled or more coarsely ground. Softer barley cultivars such as Sloop and Schooner actually had a higher GRI relative to 0.5 mmmilled samples when they were rollermilled (Figure 6). Hard barley cultivars such as Tantangara consistently had a higher GRI when they were finely ground. It should also be noted that in all cases, the GRI of roller milled barley was closer to the GRI of 0.5 mmmilled samples than the 2 mmmilled samples (Figures 5 and 6). Information of this nature could be extremely valuable to feed milling systems where savings in the operation of roller mills relative to hammer mills may also be enhanced by improvements in available energy for the target species. Extending the above example with barley, Zarrinkalam (2002) also demonstrated that increases in GRI were greatest when sorghum was 2 mmmilled relative to 0.5 mmmilled compared with wheat and barley, yet the change in GRI was the lowest when sorghum was rollermilled. Based on the suggestion by van Barneveld et al. (2001) that capacity exists to increase the energy yield from sorghum for pigs, this application of the GRI suggests that roller milling may be a far more effective way of preparing sorghum for use in nonpelleted pig diets. Conclusions A primary focus of in vitro assays in feed evaluation systems for pigs (and other livestock) has been to simulate digestion and to provide a rapid and more cost effective means of distinguishing between the nutritional quality of feed ingredients. Adoption of this approach Crude protein (%) Grimmet Mundah Galleon-N Janz-L Janz (frosted) Currawong-F Boomer-BE MR 31-B Success 42 168 van Barneveld, R.J. and Zarrinkalam, M.R. Ileal 15.00 Digestible energy 14.00 13.00 12.00 11.00 10.00 9.00 Faecal Ratio 0.90 0.88 0.86 0.82 0.80 0.78 0.76 INC99NB.3719+E Ratio 0.84 INC99NB.3719 INC99NW.1721 INCOOBS.7710 INC99NW.1721+E Grain Figure 4 Influence of exogenous enzyme application on the in vivo ileal and faecal digestible energy (MJ/kg, as received) content of wheat, barley and sorghum fed to growing pigs (R.J. van Barneveld, unpublished data). INC00BS.7710+E ef 80 70 % Change in GRI 60 50 ab 40 30 20 10 a abc bcc bc bcd bc bcde bcd bcd de de cde cde def ef f Figure 5 Change in glucose release index (%) of barley samples milled through a 2 mm screen relative to samples milled through a 0.5 mm screen. Standard error bars with different superscripts differ significantly (P<0.05) (Zarrinkalam 2002). tN er lb raGi nga F a ant Ta ngar J2 a ant Ta ngar J a nt raTa nga a nt Ta iff J Sk eron ho Sc les i rap A all dw F2 Lin eron ho Sc te et in Re ec eb Ch n-N o lle Ga on-H lle -F Ga ner hoo Sc p oo t Sl rimme G ah nd Mu B arley cultivars In vitro assessment of starch digestion in pigs 169 60 50 40 d % Change in GRI 30 20 10 0 b c c d d de d de d e e f -N ra anga J2 nt raTa nga F a nt raTa nga a nt H Ta onlle Ga c J ebe raCh nga a nt N Ta onlle Ga rt e ilb r-F2 hoonet Sc me im Gr tte e in Re les pi ra r-J A hoone Sc all ndw r-F Li hoone Sc iff Sk p oo ab a bc G Sl -10 -20 -30 a Barley cultivars Figure 6 Change in glucose release index (%) of barley samples roller_milled relative to samples milled through a 0.5 mm screen. Standard error bars with different superscripts differ significantly (P<0.05) (Zarrinkalam 2002). has been limited, largely due to a lack of agreement between the results arising from in vitro systems and in vivo measurements. In addition, other forms of analysis such as those involving near infrared spectrophotometry hold far greater potential as a means of routinely assessing the nutritional quality of grains than any method based on wet chemistry. Despite this, if we are to improve the efficiency of use of nutrients from grains for livestock we not only have to revisit the way we assess nutritional quality, focussing on more sensitive measures of energy and amino acid availability, but we have to be able to increase our capacity to assess a wider range of hypotheses associated with ways to improve nutritional quality. It is here that opportunities exist to strategically apply in vitro assays with a view to identifying those subtle changes or inherent characteristics of grains (and other feed ingredients) that may result in an improvement in livestock production efficiency. Using the glucose release index as an example, it has been demonstrated in this paper that a simple twostep enzymic procedure can be applied to identify grain components that may influence the digestion of starch in the small intestine of pigs. Further research is required to assess the extent that these influences can be observed in vivo and their subsequent effects on commercial pig production. University of Adelaide, Roseworthy SA 5371. Some of the research discussed in this paper has evolved through the Premium Grains for Livestock Program which is jointly funded by the Grains, Pig, Rural Industries (Chicken Meat and Egg Industry subcommittees) research and development corporations, Meat and Livestock Australia and Ridley Agriproducts. References Black, J.L. (2001). Variation in the nutritional value of cereals across livestock species. Proceedings of the Australian Poultry Science Symposium 13, 3039. Boisen, S. (2000). In vitro digestibility methods: History and specific approaches. In: Feed Evaluation Principles and Practice, pp. 153169 (eds. P.J. Moughan, M.W.A. Verstegen, and M.I. Visser Reyneveld). Wageningen Pers, Wageningen, The Netherlands. Englyst, H.N., Kingman, S.M., and Cummings, J.H. (1992). Classification and measurement of nutritionally important starch fractions. European Journal of Clinical Nutrition 46 (Suppl. 2), S33S50. Englyst, H.N., Veenstra, J. and Hudson, G.J. (1996). Measurement of rapidly available glucose (RAG) in plant foods: a potential in vitro predictor of glycaemic response. British Journal of Nutrition 75, 327337. Moughan, P.J. (1999). In vitro techniques for the assessment of the nutritive value of feed grains for pigs: a review. Australian Journal of Agricultural Research 50, 871881. Acknowledgements M.R. Zarrinkalam was funded by Australian Pork Limited for PhD studies at Discipline of Animal Science, 170 van Barneveld, R.J. and Zarrinkalam, M.R. van Barneveld, R.J., Nuttall, J.D., Flinn, P.C., and Osborne, B.G. (1999). Near infrared reflectance measurement of the digestible energy content of cereals for growing pigs. Journal of Near Infrared Spectroscopy 7, 17. van Barneveld, R.J., Ru, Y.J., Hughes, R.J., Flinn, P.C. and Black, J.L. (2001). Comparative digestion of energ y from grains fed to pigs, poultry and ruminants: Can the efficiency of pig production be improved? In: Manipulating Pig Production VIII, pp. 222234 (ed. P.D. Cranwell). Australasian Pig Science Association, Werribee, Victoria, Australia. Zarrinkalam, M.R. (2002). Application of a glucose release index to assess physical and chemical characteristics of cereal grains that may influence starch digestion and subsequent energy supply to monogastrics. PhD Thesis, University of Adelaide, Roseworthy, SA.