Effects of processing on bean (Phaseolus vulgaris L.) : protein quality

In animal production, feeding has an important impact on productivity and health of animals and feed composition is known to influence protein and energy metabolism directly. For monogastric animals complete diets are manufactured in which feed ingredients are used to supply the energy yielding and other nutrients. The common bean ( Phaseolus vulgaris L.) is such an ingredient.

In common beans ( Phaseolus vulgaris L.) the supply of nutrients is often lower than is expected from its chemical analysis and lower when compared to feed ingredients such as animal products. It has been shown in the present study that the utilization of nutrients from beans, particularly protein, is low due to the presence of so-called antinutritional factors (ANF). Moreover, there is a low digestibility of bean protein itself. Unless some form of processing is used, beans have little nutritional value for piglets. Various aspects of the processing of common beans and its evaluation were studied and the results reported in this thesis.

In the first chapter, a review of literature is given covering the available information on the reduction of ANF such as lectins and protease inhibitors in common beans by processing. In beans these ANF are present at either low or substantial levels. They have to be taken into account in the evaluation of bean protein quality. On the other hand, it is also the protein itself which is a component of primary interest.

It was found that the efficiency of the utilization of proteins for legume seeds depends on 1) the inherent protein resistance to proteolysis and 2) the interference by ANF, particularly the lectins.

The examination of literature revealed that research efforts so far have been directed particularly to define the processing variables for beans with respect to human consumption. This also implies the use of treatment procedures, which do not simulate the methods normally used in animal feed technology. Hence, the results of digestibility studies are difficult to compare.

It was decided to study the fractionation of beans and to investigate the distribution of ANF over the various fractions. In addition, the thermal processing of beans was studied systematically. Biological and chemical assays were also used to estimate the effectiveness of processing.

The fractionation of dry beans was studied by means of particle size reduction by fine milling and subsequent air classification. This fractionates the seed in a fines and a coarse fraction in which protein and starch are accumulated, respectively. Different procedures for milling and classification yielded fines fractions with protein levels up to 52%, being at least twice the level of the initial flour (Chapter II). The protein level in the fines fraction varied, which was mainly caused by the classifier settings under investigation. Lectins and trypsin inhibitor activity accumulated to a large extent into the fines fraction to levels being one to four-fold those of the initial flour. Expressed on a protein basis this means that air classification leads to similar or higher levels of ANF in the fines fractions. The coarse fraction also contained considerable levels of ANF. Based on the pathological effects of the fines fraction as measured in cultured explants of small intestinal mucosa of pigs it was concluded that those fractions cannot be used in feeding practice without prior elimination of at least the lectins.

For the elimination of lectins and other proteinaceous ANF, various heating techniques can be used, each of which have their advantages or disadvantages. On the basis of literature, steam processing was chosen as a principle process for systematic research into the inactivation of lectins and trypsin inhibitor activity. Therefore, a modified steam processing equipment was developed to enable high temperature/short- term (HTST) processing in addition to low temperature/long -term (LTLT) processing. The advantage of pressurized steaming (HTST) was evident in that processing times could be shortened considerably (Chapter III). This was monitored by the reduction of enzyme activity in soya beans after steaming.

The effect of heat on the inactivation of antinutritional factors and on protein quality was evaluated by steam treatment of dry beans (Chapter IV). Different time/temperature combinations were used. The inactivation of the parameters investigated can be described by first-order reaction kinetics. However, inactivation of trypsin inhibitor activity (TIA) as well as the loss of protein solubility occurred in two stages with different reaction rates, with the initial stage having a higher rate of inactivation. Lectins from beans were rapidly inactivated at temperatures of 119 and 136°C as monitored by a sensitive ELISA assay (Chapter IV). A part of the total lysine was lost on heating but the amount of available lysine was reduced to a greater extent.

All parameters investigated were highly temperature dependent. The results of this study indicate the steam processing of dry beans at 119°C for 5 to 10 min. being good enough for ANF inactivation and not too excessive for protein damage as measured by total and available lysine.

The effects of steam treatment on bean protein quality were estimated in vivo in experiments in which apparent ileal digestibility was measured in piglets fitted with a post-valvular T-caecum cannula (PVTC) (Chapter V).

In each of two experiments a control diet was used that contained casein and herring meal as protein sources. In Experiment 1 the experimental treatments consisted of beans, steam heated under atmospheric conditions (102°C) for 20, 40, 60 and 80 min., respectively. Experiment 2 comprised six experimental treatments: beans steam heated at 102°C for 60 min.; at 119°C for 5 min.; at 136°C for 1.5 min. and inclusion at 10 and 20% into the piglet diets, respectively. Intact animals were used to determine faecal digestibility of nitrogen in the same experimental treatments.

In Experiment 1, ileal digestibility of lysine and nitrogen and faecal digestibility of nitrogen were reduced significantly by the inclusion of beans, steam heated at 100°C for 20 min. Digestibility of nitrogen and lysine increased in bean containing diets with longer processing times of beans at 102°C In all cases, digestibiity coefficients of bean diets remained below the control diet.

In Experiment 2, the inclusion of 20% beans, processed at 119°C or 136°C resulted in an increased ileal digestibility of nitrogen compared to atmospheric steaming for 60 min. Steam processing, at 136°C for only 1.5 min. on the other hand, did restore ileal nitrogen digestibility to the level of the control group (Chapter V).

These studies indicate a marked difference in the biological response of piglets to diets containing beans exposed to prolonged steaming at 102°C and to HTST-processing of beans. Based on the ileal apparent nitrogen digestibility, it was concluded that HTST processing was more favourable than LTLT. The treatments of beans used in experiment 2 inactivated lectin and TIA both to a similar low level. HTST processing also resulted in a higher protein digestibility. This indicates a more pronounced role of the properties of the bean storage protein in explaining a much higher nitrogen digestibility in vivo after HTST-processing. This phenomenon requires further investigations.

Technological treatments generally involve a large number of processing variables. This also gives a large number of samples to be tested. It is important therefore to assess the potential value of alternative digestibility techniques which are more rapid to perform and which are less costly. Studies were made to evaluate the mobile nylon bag technique (MNBT) and two in vitro methods for determining and to predict the apparent ileal dry matter (DM) and nitrogen (N) digestibility (Chapter VI). A variety of steam heated bean samples was used and a comparison was made with in vivo digestibility values obtained previously by the conventional digestibility method involving total collection of ileal chyme (Chapter V).

The in vitro enzymatic procedures showed that from the procedures tested no acceptable estimates for absolute values of N and DM digestibility were obtained. Significant correlations, however, were found between MNBT ileal digestibility of N and DM and the in vivo ileal digestibility coefficients. The MNBT technique has certain advantages such as the requirement for only small amounts of feed samples and the determination of digestibility of different samples within the same animal. In this context, the MNBT appears to be promising for a rapid prediction of ileal digestibility of thermally processed feedstuffs. A disadvantage of the technique is that it requires surgically modified animals. Modification of the existing invitro techniques, therefore, is necessary, so that their prediction value for digestibility in ANF containing feedstuffs improves.