Effects of processing on antinutritional factors in legumes: the soybean case.

The author recounts his personal trail of research which has ultimately led to better understanding of the factors which contribute to the poor nutritive value of unheated soybeans. Among the techniques that were employed were the isolation of a lectin from raw soybeans, the use of affinity chromatography to remove the trypsin inhibitors, and the nutritional evaluation of soybean varieties which lacked the lectin or the Kunitz trypsin inhibitor. Based on a consideration of the results obtained by these experiments, it was estimated that the trypsin inhibitors accounted for approximately 40% of the growth inhibition on raw soy, of which two-thirds could be attributed to the Kunitz inhibitor and one-third to the Bowman-Birk inhibitor. The soybean agglutinin was deemed responsible for 50% of the inhibition of growth, and the remaining 10% is most likely due to the poor digestibility of the undenatured protein.

Possible adverse effects of soybean anticarcinogens.

For soybeans to serve as a good source of protein for feeding animals as well as humans, a certain amount of heat treatment or some other form of processing must be applied. This is because there are present in soybeans certain heat-labile factors that exert an adverse effect on the nutritional value of the protein. The so-called protease inhibitors have received the most attention in this regard and have been shown to exert their antinutritional effect in the short term by causing pancreatic hypertrophy and hyperplasia in the rat, the underlying cause for an inhibition of growth in these animals. The prolonged feeding of raw soy flour or an enriched trypsin inhibitor fraction from soybeans to rats results in the development of hyperplastic and neoplastic nodules of the pancreas, including carcinomas. It should be emphasized that all of these adverse effects are seen when protease inhibitors are present in relatively high concentrations in the diet and may be completely unrelated to the anticarcinogenic effects seen at low concentrations of the Bowman-Birk inhibitor. Brief mention is also made of any possible adverse effects that may result from the presence of phytic acid and saponins in soybeans.

Implications of antinutritional components in soybean foods.

There are a number of components present in soybeans that exert a negative impact on the nutritional quality of the protein. Among those factors that are destroyed by heat treatment are the protease inhibitors and lectins. Protease inhibitors exert their antinutritional effect by causing pancreatic hypertrophy/hyperplasia, which ultimately results in an inhibition of growth. The lectin, by virtue of its ability to bind to glycoprotein receptors on the epithelial cells lining the intestinal mucosa, inhibits growth by interfering with the absorption of nutrients. Of lesser significance are the antinutritional effects produced by relatively heat stable factors, such as goitrogens, tannins, phytoestrogens, flatus-producing oligosaccharides, phytate, and saponins. Other diverse but ill-defined factors appear to increase the requirements for vitamins A, B12, D, and E. The processing of soybeans under severe alkaline conditions leads to the formation of lysinoalanine, which has been shown to damage the kidneys of rats. This is not generally true, however, for edible soy protein that has been produced under milder alkaline conditions. Also meriting consideration is the allergenic response that may sometimes occur in humans, as well as calves and piglets, on dietary exposure to soybeans.

Effect of soybean flour on exocrine function in rat pancreas transplant with bladder drainage.

The urinary output of trypsin, chymotrypsin, elastase, and amylase by rats with a pancreas transplant and bladder drainage was determined after injection with cholecystokinin (CCK) or by feeding diets containing high (raw soy flour) or low (heated soy flour) trypsin inhibitor activity. The injection of CCK produced a significant increase in the urinary output of all four enzymes. Rats were fed heated or raw soy flour in three consecutive 10-day periods in the following sequence: period 1, heated soy flour; period 2, raw soy flour; period 3, heated soy flour. Replacing heated soy flour in period 1 with raw soy flour in period 2 caused a significant increase in the output of the four enzymes. Subsequent feeding with heated soy flour in period 3 resulted in a reduction in the output of trypsin, chymotrypsin, and elastase to levels that were not significantly different from that observed in period 1. Although amylase output was also reduced in period 3, it did not return to the level noted in period 1. These results are consistent with the roles that CCK and trypsin inhibitors are believed to play in the negative feedback control of pancreatic exocrine function. A similar approach might be employed with humans who have undergone a pancreas transplant with bladder drainage.

Effect of a trypsin inhibitor from soybeans (Bowman-Birk) on the secretory activity of the human pancreas.

This study was undertaken to clarify the issue of whether feedback regulation of pancreatic enzyme secretion occurs in humans. A naturally occurring inhibitor of trypsin and chymotrypsin, the Bowman-Birk inhibitor of soybeans, was used to reduce the activities of these enzymes normally secreted by the pancreas into the duodenum. Pure pancreatic juice was collected by endoscopic retrograde cannulation of the pancreatic duct in a protocol consisting of three periods: period 1 (15 min), collections of juice without return to the duodenum ("washout phase"); period 2 (35 min), intraduodenal infusion of juice to which buffered saline or heat-inactivated Bowman-Birk inhibitor had been added; and period 3 (55 min), intraduodenal infusion of juice in which greater than 90% of the trypsin and chymotrypsin activities had been abolished by treatment with the active inhibitor. Control experiments were included in which untreated juice was infused in period 3 as well as period 2. Enzyme analyses (trypsin, chymotrypsin, elastase, and amylase) of samples of juice collected at 5-min intervals revealed a twofold to threefold increase in the output and concentration of all four enzymes in period 3 compared with period 2. These results thus confirm the existence of feedback control in humans.

Binding of navy bean (Phaseolus vulgaris) lectin to the intestinal cells of the rat and its effect on the absorption of glucose.

The main objectives of this investigation were to study the binding of a lectin from navy beans with the epithelial cells of the rat intestine and to assess the effect of such binding on the ability of the intestine to absorb glucose. A Scatchard plot, based on the binding of 125I-labeled lectin to isolated intestinal epithelial cells, was used to calculate an association constant (Ka) of 15 x 10(6)M-1 and the number of binding sites per cell, 12 x 10(6). Metabolic studies were conducted over a period of 5 d on groups of rats fed raw or autoclaved navy bean flour and casein with or without the purified lectin. Growth, protein digestibility, biological value and net protein utilization were significantly lower in animals that had been fed raw navy bean flour or casein plus lectin than in control groups fed diets containing autoclaved navy bean flour or casein alone. Vascular perfusion was used to measure the rate of uptake of glucose by the intestines of rats that had received the various dietary treatments. The rate of absorption of [14C]glucose by intestines from rats fed raw navy bean flour or casein plus lectin was approximately one-half that of their counterparts fed the autoclaved flour or casein alone. These results provide evidence that the lectin, by virtue of its interference with intestinal absorption, is responsible, at least in part, for the nutritional inferiority of raw navy beans.

A quantitative in vitro assay of polymorphonuclear leukocyte migration through human amnion membrane utilizing 111in-oxine.

A modified amnion chemotaxis assay is described for measurement of polymorphonuclear leukocyte(s) (PMNL) migration (random and directed) into a viable membrane. The primary modifications are the use of 111In-oxine-labelled PMNL and replacement of the nitrocellulose 'trap' filter with a type I collagen sponge. The modifications resulted in four important benefits: the quantification of PMNL migration was simplified; reader subjectivity was eliminated; the information gained of the migration process was enhanced; and the assay time was decreased. The amnion chemotaxis assay with the modifications reported should provide the means of evaluating several aspects of the inflammatory response of PMNL.

A kinetic analysis of the inhibition of rat and bovine trypsins by naturally occurring protease inhibitors.

The dissociation constants (Ki) of the interaction of 10 naturally occurring inhibitors with rat anionic and bovine trypsins were determined employing three independent methods. Both enzymes bound very tightly (Ki less than 10(-9)) to bovine pancreatic, lima bean, and the Kunitz soybean inhibitors. With the exception of ovomucoid, rat trypsin bound more tightly than bovine trypsin to inhibitors derived from navy bean, lima bean, soybean (Bowman-Birk) and potato and to ovoinhibitor, leupeptin and antipain. These findings emphasize the caution that must be exercised in the interpretation of experiments involving the inhibition of trypsins from heterologous species of animals by naturally occurring inhibitors.

The failure of long-term feeding of raw soy flour, in the presence or absence of azaserine, to induce carcinogenic changes in the mouse pancreas.

The effects of feeding mice raw or heated soy flours or casein in the presence and absence of injected azaserine were investigated over a period of 18 months. Although the feeding of raw soy flour (compared with heated soy flour or casein) caused a significant inhibition of growth and an enlargement of the pancreas, there was no macroscopic evidence of pancreatic nodules in any of the six experimental groups. Microscopic examination of the pancreas revealed a somewhat higher (not significant) incidence of atypical acinar cell nodules in all animals injected with azaserine, but this difference was little influenced by the diets themselves. We concluded that raw soy flour itself has no carcinogenic effect on the mouse pancreas and does not enhance the sensitivity of the mouse pancreas to azaserine. Thus, it cannot be assumed that the appearance of pancreatic nodules constitutes an obligatory sequela of pancreatic hypertrophy and/or hyperplasia in all species of animals.

The effect of the long-term feeding of raw soy flour on the pancreas of the mouse and hamster.

The effects of the long-term feeding of mice and hamsters with raw (RSF) or heated (HSF) soy flours were investigated both in the presence and absence of chemical carcinogens. Mice which had been fed RSF for 18 months had enlarged pancreas but only a relatively low incidence of atypical acinar cell nodules (AACN). Mice on either RSF or HSF were also relatively resistant to the carcinogenic effects of azaserine which is known to produce a high incidence of AACN in rats. Hamsters which had been fed RSF for 15 months did not exhibit pancreatic enlargement and had a very low incidence of pancreatic tumors. Although the incidence of tumors in hamsters which had been injected with N-nitrosobis (2-oxopropyl)amine and maintained on HSF was very high (88%), those on RSF had a tumor incidence of less than 10%. Thus, there appears to be a marked difference in the response of the pancreas in different species of animals to the long-term effects of feeding RSF. This should be taken into consideration in evaluating the potential carcinogenicity of the trypsin inhibitors.

Effect of long-term feeding of soy-based diets on the pancreas of Cebus monkeys.

Feeding soy-based protein containing trypsin inhibitor causes pancreatic hypertrophy in the rat, and long-term feeding (up to 2 years) has revealed a high incidence of adenoma following hypertrophy. It was therefore of interest to determine whether the ingestion of soy-based protein has any adverse effects on the primate pancreas. A resource of 27 Cebus albifrons monkeys, previously used to evaluate the protein quality of several soy and milk proteins, has been maintained on semi-synthetic diets for 3 to 4 years; the protein sources for the diets were casein, lactalbumin, soy isolate and soy concentrate. In general the monkeys were in good physical health and their weights were appropriate for age and sex. Serum biochemical and hematological profiles were normal and there were no major differences between the groups. A pancreatic biopsy from both the head and tail region of the pancreas was taken from each monkey. Visual observation of the pancreas revealed no overt pathology; two independent histological examinations indicated no diet-related differences between groups, and biochemical analyses of trypsin, chymotrypsin, protein, DNA and RNA revealed no differences. It is concluded that feeding low level trypsin inhibitor-containing diets for up to 4 years caused no adverse effects in the pancreas of the Cebus nonhuman primate.

Potentially defensive proteins in mature seeds of 59 species of tropical leguminosae.

A survey of 59 species of tropical legume seeds revealed high interspecific variation in proteinaceous capacity to inhibit bovine trypsin (a digestive enzyme) and to agglutinate human (type B, Rh positive) and laboratory rabbit red blood cells. The legume subfamily Mimosoideae was conspicuous for the absence of seeds with very weak trypsin inhibition. Congenerics sometimes differed strongly from each other with respect to both trypsin inhibition and phytohemagglutination. Half the species of seeds displayed no hemagglutinating capacity with one or the other kinds of red blood cells, and in only 27% of the 30 cases where there was some activity did the same species of seed actively agglutinate both species of red blood cells. A species of seed that had hemagglutinating capacity was almost invariably associated with moderate to high levels of trypsin inactivation. While it has been long known that a great diversity of small toxic and potentially defensive molecules occur in legume seeds and that one species of seed often contains several of them, we now feel that it is reasonable to consider legume seeds as also containing a high diversity of potentially toxic protein molecules. A single seed is likely to contain, at the least, three to four classes of defensive compounds, any or all of which, or some in combination, may be the cause of a seed being rejected by a potential seed predator.

The effects of long-term soy protein and milk protein feeding on the pancreas of Cebus albifrons monkeys.

Twenty-seven 2- to 4-yr-old cebus monkeys (Cebus albifrons) were fed from infancy purified diets containing lactalbumin, soy isolate, casein or soy concentrate as the sole protein source. Hematologic and clinical chemistry values were similar for all groups. Head and tail portions of each pancreas were surgically removed for histopathologic evaluation and determination of protein, RNA and DNA content, and for trypsin and chymotrypsin activity. Hematoxylin and eosin-stained sections from 26 of 27 monkeys showed normal pancreatic tissue with occasional acinar vacuolation in all diet groups. The remaining animal, one of only two fed soy concentrate, had diffuse interstitial fibrosis of the pancreas associated with mild to moderate atrophy of acinar tissue. Biochemical analyses of the pancreatic biopsies indicated no group differences among animals fed lactalbumin, soy isolate or casein. One of two monkeys in the soy concentrate group showed decreased pancreatic protein, RNA and trypsin concentrations; this was probably due to the fibrosis in this animal. No evidence of pancreatic hypertrophy or hyperplasia, as measured by RNA/DNA and protein/DNA ratios, respectively, was seen in any diet group.

Protection against pulmonary oxygen toxicity in rats by the intratracheal administration of liposome-encapsulated superoxide dismutase or catalase.

To test the feasibility of using liposomes to deliver therapeutic agents to the lungs, the effect of liposome-encapsulated superoxide dismutase (SOD) or catalase on pulmonary oxygen toxicity was studied in rats. The SOD or catalase was encapsulated in negatively changed multilamellar liposomes and administered directly into the trachea of adult rats, which were subsequently exposed to hyperoxia (greater than 95% O2). Response to hyperoxia was examined by studying lung SOD and catalase activities, survival rates, and lung morphology. Rats receiving liposome-encapsulated SOD or catalase showed increased levels of enzyme activities in the lung homogenates compared with those in the control groups after 24 to 72 h of hyperoxic exposure. Elevated enzyme levels in the lungs of rats treated with liposome-encapsulated SOD or catalase were accompanied by a significant improvement in survival rates after 72 h of hyperoxic exposure and less lung injury than in the other control groups.

The prevention of elastase-induced emphysema in hamsters by the intratracheal administration of a synthetic elastase inhibitor bound to albumin microspheres.

The peptide, succinyl-alanyl-alanyl-prolyl-valine chloromethylketone (SPCK), a synthetic inhibitor of elastase, was covalently attached to human albumin microspheres (HAM) and administered intratracheally to hamsters 15 min and 8 h prior to the instillation of porcine pancreatic elastase. Pressure-volume relationships and histologic examination of excised lungs after 4 wk showed complete protection from emphysema when SPCK-HAM was administered either 15 min or 8 h before elastase exposure. Concurrent experiments with free SPCK showed that protection was achieved only if elastase was administered within 15 min after the instillation of SPCK. Extending this period to 8 h not only led to a failure of free SPCK to prevent emphysema but actually resulted in more extensive air-space enlargement. The prolonged effectiveness conferred by the attachment of SPCK to a biodegradable carrier should reduce the frequency with which it would have to be administered for the therapeutic intervention of emphysema and should minimize any toxic side reactions.

Effect of raw soy flour, feeding regime, and azaserine on rat pancreas.

The combined effects of feeding rats increasing amounts of raw soy four, feeding regime (ad libitum vs one meal per day), and injection of azaserine on the incidence of pancreatic nodules were investigated over a period of 12 months. Food consumption and body weights of meal-fed rats were lower than those of their ad libitum counterparts. The difference in body weight between the ad libitum and meal-fed rats became greater as the level of raw soy flour in the diet increased. Azaserine injections did not affect food consumption or body weight. The weights of the pancreas (gm/100 gm BW) increased parallel to the level of raw soy flour in the diet. The survival rate of rats on diets containing 19% and 42% raw soy flour was greater on the meal-fed regime than in the ad libitum group, but the converse was true when the diet contained 80% raw soy flour. The number and severity of pancreatic nodules observed in azaserine-injected animals surviving at the end of 12 months increased in relation to the level of raw soy flour in the diet and was enhanced at each level by meal feeding.

The effects of soy flour and N-nitrosobis (2-oxopropyl) amine on the pancreas of the hamster.

Syrian golden hamsters fed diets of raw or heated soy flour for 16 months had an incidence (4%) of pancreatic neoplasms in both groups. Animals injected with N-nitrosobis (2-oxopropyl) amine (BOP) and fed heated soy flour for 15 months developed a high incidence (88%) of microscopic benign and malignant neoplasms, primarily of ductal origin. This was in marked contrast to a similar group of BOP-injected animals which had been fed raw soy flour and in which the incidence of pancreatic neoplasms was less than 10%.